Workbook 3 - apps.who.intthe general steps for evaluation de-scribed in Workbooks 1 and 2. Use this specialised workbook simul-taneously with the foundation workbooks to maximise the

1Workbook 3 · Needs Assessments

WHO/MSD/MSB 00.2d

Workbook 3

NeedsAssessment

2 Evaluation of Psychoactive Substance Use Disorder Treatment

WHO/MSD/MSB 00.2d

WHOWorld Health Organization

UNDCPUnited Nations International Drug Control Programme

EMCDDA European Monitoring Center on Drugs and Drug Addiction

World Health Organization, 2000c

This document is not a formal publication of the World Health Organization (WHO) and all rights are reserved by theOrganization. The document may, however, be freely reviewed, abstracted, reproduced and translated, in part or inwhole but not for sale nor for use in conjunction with commercial purposes. The views expressed in documents bynamed authors are solely the responsibility of those authors.


WHO/MSD/MSB 00.2d

ited the workbook series in laterstages. Munira Lalji (WHO, Sub-stance Abuse Department) and Jen-nifer Hillebrand (WHO, SubstanceAbuse Department) also edited theworkbook series in later stages.Maristela Monteiro (WHO, Sub-stance Abuse Department) pro-vided editorial input throughout thedevelopment of this workbook.

Some of the material in this work-book was adapted from a NIDApublication entitled “How Good isYour Drug Abuse Treatment Pro-gram? A Guide to Evaluation.” Con-tributions drawing from this reportare gratefully acknowledged.

The World Health Organizationgratefully acknowledges the contri-butions of the numerous individu-als involved in the preparation ofthis workbook series, including theexperts who provided useful com-ments throughout its preparation forthe Substance Abuse Department,directed by Dr. Mary Jansen. Finan-cial assistance was provided byUNDCP/EMCDDA/Swiss FederalOffice of Public Health. Cam Wild(Canada) wrote the original text forthis workbook and Brian Rush(Canada) edited the workbook se-ries in earlier stages. JoAnneEpping-Jordan (Switzerland) wrotefurther text modifications and ed-

Acknowledgements


WHO/MSD/MSB 00.2d


WHO/MSD/MSB 00.2d

Table of contents

Overview of workbook series 6

What is a needs assessment? 7

Why do a needs assessment? 7

How to do a needs assessment? 8

Question 1 10

Question 2 14

Question 3 16

Question 4 23

Comments about case examples 28

Case example of a needs assessment 29Planning and evaluating outpatient care for drugdependent patients in Barcelona (Spain)

Case example of a needs assessment 42A study to determine the welfare service needsin the Eastern Transvaal, Republic of South Africa


WHO/MSD/MSB 00.2d

Overview ofworkbook seriesThis workbook is part of a series in-tended to educate programme plan-ners, managers, staff and other deci-sion-makers about the evaluation ofservices and systems for the treat-ment of psychoactive substance usedisorders. The objective of this se-ries is to enhance their capacity forcarrying out evaluation activities.The broader goal of the workbooksis to enhance treatment efficiency

and cost-effectiveness using the in-formation that comes from theseevaluation activities.

This workbook (Workbook 2) de-scribes step-by-step methods forimplementing evaluations. Thesesteps span from starting the study, tocollecting, analysing, and reportingthe data, to putting the results intoaction in your treatment programme.

Introductory WorkbookFramework Workbook

Foundation WorkbooksWorkbook 1: Planning EvaluationsWorkbook 2: Implementing Evaluations

Specialised WorkbooksWorkbook 3: Needs Assessment EvaluationsWorkbook 4: Process EvaluationsWorkbook 5: Cost EvaluationsWorkbook 6: Client Satisfaction EvaluationsWorkbook 7: Outcome EvaluationsWorkbook 8: Economic Evaluations


WHO/MSD/MSB 00.2d

What is a needsassesment?

Needs assessment is a toolfor program planning.

Over the last two decades, the roleof needs assessment in the planningof services and systems for PSU dis-orders has increased in importance.Several factors have contributed tothis development, including:

Questions that arise about the rela-tive priority of different commu-nity needs. In some jurisdictionswith no services for PSU disor-ders, the focus is now to ask aboutnew services that might be

needed. In other areas with avail-able services, the focus is now toask about how existing servicesmight be better co-ordinated andmore efficient.

The increasing diversity of com-munity interventions that are avail-able. There is acceptance in mostjurisdictions that a range of com-munity services is needed and thatpeople coming into treatmentshould be appropriately assessed

Why do a needsassesment?

Needs assessments evaluate:

The capacity of treatment servicesin the community in relation to theprevalence and incidence of PSUdisorders

The appropriate mix of servicesrequired to respond to the di-

verse needs associated with PSUdisorders

The co-ordination of serviceswithin a system of care in order tofacilitate entry into the system,smooth transition across specificcomponents and appropriate fol-low-up

l

l

l

l

l


WHO/MSD/MSB 00.2d

and matched to treatment. How-ever, information is needed to helpdecide how much of what type oftreatment is required in a givencommunity or region.

The increasing use of PS with po-tential for harm among the generalpopulation, and among peopleseeking treatment.

The desire to take a more preven-tive approach to PSU disordersand to improve the balance oftreatment, early intervention and

prevention-oriented programmesin the community.

In any case, the specific objectivesof the needs assessment must beclarified and documented. Theseobjectives may include:

to respond to an external mandatefor needs assessment prior to ap-proval and release of funds

to guide the allocation of newfunding among several new op-tions being considered

Most experts in the field of PSU dis-orders agree that a single "all-pur-pose" needs assessment techniquedoes not exist. This is because needsassessment planners have differentgoals for conducting assessmentsmaking it unlikely that a singlemethod would suffice for all pur-poses.

In this workbook, various ap-proaches to needs assessment aredescribed by showing how they canbe used to address four questionsmost commonly asked in a needsassessment project. More details re-garding many of these approachescan be found in recent reviews

How to do a needsassesment?

In thisworkbook,variousapproachesto needsassessmentare described byshowing howthey can be usedto address fourquestionsmost commonlyasked ina needsassessmentproject.

(DeWit and Rush, 1996). The fourquestions addressed are:

1 How many people in the regionor community need treatment forPSU disorders?

2 What is the relative need for treat-ment services across different re-gions or communities?

3 What types of services are neededand what is the necessary capac-ity?

4 Are existing services co-ordinatedand what is needed to improve theoverall level of system functioning?

l

l

l

l


WHO/MSD/MSB 00.2d

Use thisspecialisedworkbooktogether,simultaneouslywith thefoundationworkbooks tomaximise theinformation thatis presented.

The two case examples at the end ofthis workbook present two very dif-ferent approaches to needs assess-ment. The first (from Spain) reliesupon existing computerised data-bases, whereas the second (fromSouth Africa) uses interviews andfocus groups. Despite their differ-ences, both evaluations are appro-priate because they take into accountthe unique needs and resources oftheir settings.

Each of these questions, and themethods for answering them, areaddressed below. Keep in mindthat this information is supplemen-

tary to the general steps for evalu-ation outlined in Workbooks 1 and2. When doing a needs assessment,you should carry through each ofthe general steps for evaluation de-scribed in Workbooks 1 and 2. Usethis specialised workbook simul-taneously with the foundationworkbooks to maximise the infor-mation that is presented.

Using Workbook 1 as a guide, de-termine which one of the above fourquestions is most relevant for yourprogramme evaluation question.Review that section below.


WHO/MSD/MSB 00.2d

Question 1

1. Mortality-basedprevalence models

This method is easy to use, if you have thenecessary data. For alcohol, for example,the formula is:

A = P*(D/K),

where

A = the total number of alcohol dependent persons inan area or region

P = the proportion of liver cirrhosis deaths due to alcohol use

How many people in theregion or communityneed treatmentfor PSU disorders?This workbook will briefly describethree approaches to answering thisquestion. Unfortunately, there is noeasy answer to this question because thevarious strategies available to you each

come with their unique advantages anddisadvantages. The selection will haveto depend on your unique circum-stances and the expertise, time and re-sources that are available.

O = the total number of deathsfrom liver cirrhosis reportedfor a given year in the areaor region of interest

K = the annual death rate fromliver cirrhosis among alcohol dependent personswith complications (e.g.,rate of death from livercirrhosis per 10,000 alcoholdependent persons).


WHO/MSD/MSB 00.2d

By collecting the necessary statistical in-formation for a region or community, oneshould be able to fill in the required infor-mation and estimate the number of prob-lem alcohol users. This is used as the esti-mate of the number of people in need oftreatment.

Advantages include:

l The simplicity of the formula, once thenecessary statistical data are obtained

Limitations include:

l inaccuracies in the statistical data due tomisclassification of the cause of death

l instability of the prevalence estimatesfor small populations since deaths due

to liver cirrhosis (or suicide and alco-hol use) occur infrequently

l the need to supplement the resultingestimates of the in-need populationwith estimates based on PS other thanalcohol

l limited utility from prevention or earlyintervention perspectives because esti-mates are based on the most severe con-sequences of alcohol use

l variations in the constants in the for-mula across cultural and social settings

l inability to estimate the number ofpeople in need of treatment within spe-cific population sub-groups (e.g., gen-der, age)

2. General population survey

In a general population survey, you con-tact a random or representative sampleof people in the region or communityand ask them questions about theirPSU, related problems and perceivedneed for treatment. A survey of thistype can be easy or difficult to com-plete, depending on the complexity ofyour evaluation.

It is important to pay attention to thequestions you ask and to the criteria youuse to indicate whether the respondentA ”needs” treatment for a PSU disorder.You may choose to ask questions aboutthe amount and pattern of drinking overa recent time period (e.g., to calculate av-erage weekly consumption, or the num-ber of respondents drinking more than acertain number of drinks on a given day).The main limitation of these data is theunder reporting of consumption that usu-

ally occurs in a population survey; oftenunderestimating actual consumption by asmuch as 50%-60%. Alternatively, youmay ask questions about problems theperson has experienced related to theirPSU and create a cut-off point on the listof problems to define the need for treat-ment. Many people conducting a popu-lation survey create their own problemlist but this raises significant questionsabout the reliability and validity of thesurvey items.

Various survey instruments have been de-veloped that are appropriate for use in ei-ther face-to-face or telephone interviews.An excellent example for use in many cul-tural settings is the Composite Interna-tional Diagnostic Interview (CIDI) devel-oped by the World Health Organization(Cottler et al., 1991; Robins et al., 1988;Wittchen et al., 1991; WHO, 1990). The


WHO/MSD/MSB 00.2d

basic approach is to establish the presenceor absence of a set of A “symptoms”,which include both clinical manifestations(e.g., tolerance, withdrawal, craving) andsocial consequences (e.g.,PSU-relatedproblems with family, friends, job, and/orthe criminal justice system). To be assigneda particular diagnosis, an individual mustmeet predetermined counts of such “symp-tom”.

Advantages include:

l direct estimates of the number of peoplein need of treatment for PSU disorders.

Limitations Include:

l important segments of the populationare difficult to reach in a population sur-vey either because they are hard to lo-cate (e.g., homeless) or because theyare excluded in the sampling procedure

(e.g., incarcerated, institutionalised).Other strategies will be needed to esti-mate the in-need treatment populationin these groups

l results of surveys may be biased if theresponse rate is lower for particularsub-groups such as young adults, theelderly, women or particular cultural/ethnic groups

l there is a heavy reliance on the respon-dents’ self-report of consumption, re-lated problems, and there will be a gen-eral tendency to underestimate PSU andrelated problems

l some survey methods are very expen-sive (e.g., face-to-face interviews) andrequire special expertise that may needto be purchased on a consulting basis ifit is available (e.g., survey statistician,trained interviewers, data analyst)

3. Capture-recapture models

This method requires that you have accessto computerised records and a certain levelof statistical expertise. Its advantage is thatit overcomes the difficulty of accessinghard-to-reach segments of the PSU popu-lation by relying on sources of informa-tion that contain “naturalistic” samples ofknown PS users. These sources of infor-mation might include police records ofarrest for possession of narcotics or courtconvictions for PSU-related crime, hospi-tal emergency room admissions involvingcases of PS overdose or admissions to PSUtreatment centres. Used in isolation, thesedata sources are not particularly helpfulfor estimating prevalence. However, com-bining data from two or more sources ofinformation can yield reliable and valid

The term�capture -recapture� isderived from thisprocess in whichindividuals in thefirst sample orlist are capturedand identified(tagged), andthen a certainportion are re-captured or re-identified on thesecond list.

estimates of the total population of PSusers.

The case example from Spain, located atthe end of this workbook, uses the cap-ture-recapture method for a portion of itsanalyses. Their data sources includedrecords for treatment admissions, emer-gency visits, and jail entrances.

The logic of the capture-recapture modelfor estimating hidden populations of PSusers is best understood by way of an ex-ample. Suppose that for a given area orregion, one has two separate listings ornaturalistic samples of known opioid us-ers. The first list, which we will call list X(sample 1), consists of opioid-related ar-


WHO/MSD/MSB 00.2d

rest cases and the second list called list Y(sample 2), consists of opioid overdosecases presenting to hospital emergencyrooms. With two lists or samples, thereare four possible locations where any givenindividual may appear: on list X and noton list Y, on list Y and not on list X, on listX and on list Y and finally on neither listX or list Y. Figure 1 presents the range ofpossible locations in the form of a contin-gency table.

In the figure on the next page, the onlyunknown is cell f22, the frequency countof the number of cases appearing on ei-ther list or sample. Once we obtain thenumber of cases appearing in the first threecells, it becomes possible to estimate cellf22, and subsequently the total populationof opioid users.

Obtaining a value for the first cell (f11)requires that researchers attach uniqueidentifiers to each case appearing on bothlists. Examples of unique identifiers in-clude date of birth, gender, marital statusor ethnicity. Once this procedure is com-plete, it becomes possible to match thenumber of individuals or cases appearingon both lists. The term “capture-recap-ture” is derived from this process in thatindividuals in the first sample or list arecaptured and identified (tagged), and thena certain portion are re-captured or re-identified on the second list. The largerthe number of unique identifiers, thegreater the precision in matching cases.Cells f12 and f21 are easily estimated us-ing the same identifying procedures. Withvalues for the first three cells determined,the following formula, known as thePeterson estimator, may be used to esti-mate cell f22:

With values for the first three cells deter-mined, the following formula, known asthe Peterson estimator, may be used to es-timate cell f22:

With f22, an estimate of the total popula-tion of opioid users is given by:

There is no restriction on the number oflists (samples) that may be used in the cal-culation of the estimate. In fact, the greaterthe number of independent listings orsamples of opioid users, the more accuratethe estimate becomes.

Yes No

Yes f11 f12

No f21 f22=?

Advantages include:

l a low-cost approach for helping to esti-mate the number of people in need oftreatment for PSU disorders in your re-gion or community.

Disadvantages include:

l potential violation of the assumptionsunderlying the model, for example, in-dependence of the samples (i.e., beingon one list doesn’t influence the prob-ability of being on the other)

l contamination of the samples through at-trition (e.g., death) or mis-classification

l the length of time required to clean thelists and match cases if the unique iden-tifiers

l lack the required detail and specificity

l limited background information about thePS users on the lists making it difficult todetermine the types of treatment servicesthat may be most appropriate for them

Case in list X(sample 1)

Case in list Y(sample 2)


WHO/MSD/MSB 00.2d

Question 2What is the relative needfor treatment servicesacross different regionsor communities?

One way to answer this question is to com-pare the prevalence of the in-need treat-ment population as established with oneof the three methods described in the abovesection. However, other, more easily ob-tained statistical data may also be avail-able that are correlated with PSU disor-ders in the community. Geographic areascan then be ranked on the various indica-tors and then all the indicators combinedinto one index that reflects PSU disorders.The index may then be used to comparethe relative level of these disorders acrossthe regions. This method requires that youhave access to computerised records andthat you have the resources and expertiseto perform computer-based statisticalanalyses.

Examples of indicators include indices ofalcohol availability (e.g., number of liquorstores per 100,000 population); mortality(e.g., rate of alcohol-related deaths per

100,000 population); poverty (e.g., per-cent owner-occupied units with water sup-ply and/or electricity), and drunk drivingand traffic accidents (e.g., rate of driversinvolved in personal injury accidents by100,000 licensed drivers).

Once the individual indicators have beenselected, you have different options forcombining them into an overall index. Fairlysophisticated statistical procedures such ascluster analysis and factor analysis havebeen used to create this index (Beshai,1984; Tweed and Ciarlo, 1992; Tweed etal., 1992). Adrian (1983) presents two lesscomplicated methods. The first approachinvolves ranking each indicator across thevarious geographic areas being compared.A mean rank is then calculated for eachindicator and the mean rank for the indica-tor is then ranked across the areas into anoverall rank. This approach weights eachindicator equally and has the advantage of

�Questionsabout therelative needfor services forPSU disorderscan beanswered withindices thatcombineinformation onseveralproblemsrelated to thenature andprevalence ofthesedisorders.�


WHO/MSD/MSB 00.2d

being easy to calculate and interpret. Thedisadvantage is that the approach is rela-tively insensitive to the magnitude of thedifference between ranks.

The second approach used by Adrian(1983) first gives a value of 100 to theoverall rate for each indicator, for all ar-eas combined. The small area rates arethen calculated as a fraction relative tothe overall rate. For each area, the meanof the various indices is then calculatedto create the composite PSU index. Un-like the ranking method, this index ap-proach is sensitive to the degree of dif-ference in the ranks between the areasbeing compared. The main disadvantageis that the mean of the individual indicesis sensitive to extremely high values. Theindex method is more helpful in assess-ing relative need because it retains thedegree of difference across the areasbeing compared, and thus the relativeimportance of different indicators. Amap of the different areas being com-pared can also be developed showing thevariation in the level of PSU disordersin relation to the average for the entireregion.

The main limitation of all these approachesto comparing different areas on the rela-

tive need for services for PSU disorders isthe reliability and validity of each of theindividual indicators. For example, manysocial indicators (e.g., income level, hous-ing) have only indirect relationships to PS.Other indicators, such as drunk drivingarrests and convictions, will be influencedby the level of policing and judicial discre-tion. While it can be argued that the dis-advantages of one indicator can be offsetby the advantages of another, indicatorsshould only be selected if they are reliable,valid and of comparable meaning acrossthe regions.

In summary, questions about the relativeneed for services for PSU disorders canbe answered with indices that combine in-formation on several problems related tothe nature and prevalence of these disor-ders. After one has compared a region orcommunity to other areas a stronger ar-gument for reallocating resources may bepossible. However, neither the estimatesof the in-need population, nor the relativeneed for services compared to other ar-eas, provide much direction in determin-ing the type of services or the amount ofthese services that are needed. Other needassessment strategies are required to an-swer such questions and these are de-scribed below.


WHO/MSD/MSB 00.2d

Client-centred Community Needs Assess-ment (CCCNA) is a prospective data col-lection procedure that assesses what cli-ents or patients think about services thatare needed. It has been applied in bothmental health (Cox et al., 1979) and sub-stance use treatment services (DiVillaer,1990 & 1996). It is easy to complete, andhas the added advantage of assessing thepoint-of-view of potential consumers ofprogramme services. There are four im-portant assumptions underlying this ap-proach:

l community needs should be identified,at least in part, on the basis of compre-hensive clinical assessment of a largeand representative sample of individu-als in need

Question 3

What types of servicesare needed and what isthe necessary capacity?

Client-centered CommunityNeeds Assessment

l needs should be expressed as specific typesof intervention (e.g., outpatient PSU disor-der counselling; life skills training) that canbe established in the community

l relevant demographic and clinical in-formation on those individuals in needof the interventions should be collected

l there should be some assurance that thoseindividuals in need of the interventionswould actually use the interventions ifestablished in the community

This method asks about basic client infor-mation (e.g., gender, age), his/her PSUbehaviour, and information about the“ideal” intervention required by the client. The listed intervention is then coded as:


WHO/MSD/MSB 00.2d

1 the intervention does not exist in the com-munity

2 the intervention exists but is not available(i.e., agency admission criteria rule out thisclient) or accessible (i.e., certain factorssuch as transportation, hours of operationrule out participation)

3 the intervention exists and is available and/or accessible to the client

4 the intervention exists and is availableand/or accessible to the client, but theclient is unwilling to attend the agencythat offers it in the community

As the information accumulates about the sta-tus of interventions needed for particular typesof clients, a profile emerges of important gapsin service in the community on region.

The main advantage of this needs assess-ment strategy is that it incorporates in-formation directly about the person in need,

This approach is easy to complete and doesn’trequire sophisticated computer-based analyses.In this approach, you list the types of PSU dis-order services that ideally should be available topeople in a region or community, and then con-trast this ideal template with the actual state ofaffairs. Although there is no international stan-dard for the list of various services that shouldbe used as a template, there is wide agreementthat the ideal treatment system should reflect a“continuum of care.” The rationale underlyingthe continuum of care is that the population inneed of treatment for PSU disorders is highlyvaried and that many different types of servicesis needed to meet these diverse needs. Com-

as well as the client’s own perception ofthe suitability of different service options tomeet their needs. However, the CCCNAmethod is limited in the following ways:

l needs of people presenting for treat-ment may not reflect the needs of allpeople experiencing disorders in thecommunity

l the lack of widely agreed upon criteriafor matching clients to treatment meansthat considerable judgement is in-volved on the part of clinicians and cli-ents in establishing the “ideal” treat-ment intervention

l depending on the number of agenciesinvolved, considerable time and re-sources may need to be dedicated totraining of personnel, monitoring thequality of the data collection andanalysing and reporting the resultinginformation

Continuum of care approach

prehensive assessment and matching to treat-ment ensures effective use of each type of ser-vice in the treatment system.Your list of ser-vice types might include:

l case identification

l comprehensive assessment

l case management

l withdrawal management (home/facility)

l brief intervention

The rationaleunderlying thecontinuum ofcare is that thepopulation inneed oftreatment forPSU disordersis highly variedand that manydifferent typesof services isneeded to meetthese diverseneeds.


WHO/MSD/MSB 00.2d

l methadone maintenance

l outpatient counselling

l day/evening treatment

l short-term inpatient treatment

l long-term impatient treatment

l supportive housing

l continuing care

l mutual aid

With your list completed, and appropriatedefinitions developed, you then examine theservices in each region or community underinvestigation and determine whether the ser-vice is:

l available; that is whether participa-tion in the service is restricted by cer-tain admission criteria (e.g., no legalchanges pending; must be male only);

l accessible; that is whether factorsmake use of the program difficult (e.g.,lack of public transportation, hours ofservices, language of service provi-sion).

Template to assess the availability and accessibility of servicesalong an ideal continuum of care

Region #2 Region #3Region #1 Region #4

Case identification

Comprehensive assessment

Case Management

Withdrawal/ Mgmt - (home)

Withdrawal/ Mgmt - (social)

Withdrawal/ Mgmt - (facility)

Methadone Maintenance

Brief Intervention

Outpatient Counseling

Day/ evening treatment

Short-term Inpatient treatment (medical)

Short-term Inpatient treatment (non medical)

Long-term Inpatient treatment

Supportive Housing

Continuing Care

Mutual Aid

Avail Acc Avail Acc Avail Acc Avail Acc

youthonly

female only

cocaine usesonly


WHO/MSD/MSB 00.2d

The information about availability and ac-cessibility may come from a formal surveyof service providers, a review of previousservice inventories or interviews with localkey informants.A simple check-off proce-dure will provide a crude overview of thegaps in the treatment system in terms of theavailability and accessibility of specific typesof services. One should also provide a briefwritten description of service availability andaccessibility. The table on the previous pagemay serve as a template for those adoptingthis approach.

Advantages include:

l easy to conceptualise and implement

l allows for creative thinking about newservice options not previously adoptedin the region(s)

Normative need assessment models areessentially “demand-based”, that is pro-jecting future needs on the basis of pastdemand on, and performance of, thetreatment system. This approach is fairlycomplicated, and best for those with com-puter and statistical resources. The mostsophisticated of these approaches alsotakes into account local variation in theprofile of PSU disorders.

The Alcohol Treatment Profile System(ATPS) developed in the U.S.A. is agood example of a normative needsassessment model (Ryan, 1984/1985).The ATPS has two main components.The first component, referred to as the“need” component, was developedbased on seven mortality-based indi-cators reported as average annualdeath rates per 100,000 population forthe age group 15 to 74, and for theperiod 1975-1977. The indicators in-


l lack of standardisation across jurisdic-tions concerning the components of anideal treatment system and definitions ofthe service categories

l insufficient attention to the quality ofexisting services and the evidence re-garding specific types of treatment in-terventions that they offer (e.g., cog-nitive-behavioural vs. drug therapy)

l insufficient attention to the flow of cli-ents across these service componentsand other dimensions of system-levelco-ordination

l inability to quantify the required capac-ity and resource complement (e.g.,staff, beds) of the services consideredto be needed

Normative approach

cluded death rates from liver cirrho-sis, alcohol dependence, alcohol poi-soning, suicide, homicides, automobileaccidents and alcohol-related psycho-sis. These indicators were factor-analysed and two separate indices ofalcohol-related problems emerged.The first factor was called a ChronicHealth Index and was used to estimatethe prevalence of chronic, long-termalcohol-related problems. The secondfactor, called the Alcohol Causality In-dex, was used to estimate the preva-lence of acute alcohol intoxication.The value of this index does not indi-cate how many individuals suffer fromacute intoxication or chronic long-termproblems, but rather indicates “relative”prevalence ratings for individual coun-ties. The mortality indicators are avail-able nationally at the county level. Con-sequently, an index value for each countyhas been calculated and published.


WHO/MSD/MSB 00.2d

The second component of the ATPS nor-mative model is the “demand” componentand is based on treatment data collectedat the national level by the National Drugand Alcohol Treatment Survey(NDATUS) (Harris & Colliver, 1989). Thesurvey data provide estimates of the levelsand patterns of existing service use and ser-vice capacities for each planning area acrossthe country. Level of use is expressed asthe number of clients served. Service ca-pacity is expressed as the number of treat-ment slots. NDATUS classifies treatmentinto seven different modalities: medicaldetoxification, social detoxification, rehabili-tation, custodial, ambulatory, limited careand outpatient. Service use and capacity areestimated separately for each of these treat-ment modalities.

In the ATPS model, the NDATUS dataform the dependent variable. “Observed”treatment service levels and capacities foran area are therefore modelled as a func-tion of the two indices of alcohol-relatedproblems. Because the relationship be-tween need and demand varies substan-tially according to different populationsizes, population size is included as a thirdindependent variable in the model. Esti-mates of total expected clients and totaltreatment capacities and estimates bro-ken down by treatment modality, are re-lated to an area’s Chronic Health Index,its Alcohol Causality Index and its popu-lation size. For planning purposes, esti-mates of expected clients and treatmentcapacities are presented in a series oftables according to an area’s populationsize, Chronic Health Index and AlcoholCausality Index. Needs assessment plan-

ners use these tables to compare the ex-pected treatment capacity of a county im-plied by the normative model with thecounty’s actual or observed capacity.

Advantages include:

l ease of use once the necessary informa-tion has been compiled

l for each estimate, the model providesa high and low range for a given plan-ning area and this is helpful in applyingthe results in the decision-making pro-cess


l the social and health indicators thatcomprise the problem indices in themodel are subject to a wide variety ofbiases

l the data on past treatment serviceutilisation may not be based on allexisting treatment facilities sincesome may not have participated inthe survey or otherwise have beenexcluded (e.g., treatment in the pri-vate sector)

l the assumption that current or pasttreatment service utilisation patternsare an adequate reflection of currentclient needs at the time services areprovided and in the near future. Forexample, the needs of the potentialpopulation of service users may notbe identical to the needs of the clientpopulation who have sought treat-ment in the past


WHO/MSD/MSB 00.2d

Unlike the demand-based ATPS normativemodel, which relies on what actually exists inthe treatment system in terms of serviceutilisation patterns, prescriptive modelsspecify the level of treatment services thatshould or “ought” to be provided to the resi-dents of a given planning region. This ap-proach can be seen as an extension of the“continuum of care” approach describedabove, but more complicated and requiringmore computer resources.

Prescriptive models usually begin with aprevalence estimate of the size of the popu-lation in need. It is realistic to assume thatnot all of these individuals will voluntarilyseek treatment and that there are only lim-ited resources available to treat those whocome to the attention of treatment special-ists. An objective, then, is to determine whatproportion of the in-need population shouldreceive treatment in a given year. Many pre-scriptive models arrive at a figure of 20%based on a series of “assumptive” values orproportions assigned to the population withalcohol use disorders in a region or area (seebelow). This final value, indicating the levelof “demand” for treatment services, is thenapportioned throughout various componentsof the ideal treatment system (detoxification,case management, etc.).

Ford (1985) describes a standard set ofprocedures to arrive at the 20% estimateof the proportion of the in-need popula-tion to be treated each year:

1 Two-thirds of alcohol dependent per-sons drink again within one year oftreatment.

2 The rate of increase in alcohol depen-dence is around 10 percent per year.

Unlike demand-based normativemodels, whichrely on whatactually exists inthe treatmentsystem in termsof serviceutilisationpatterns,prescriptivemodels specifythe level oftreatmentservices thatshould or�ought� to beprovided to theresidents of agiven planningregion.

Prescriptive approach

3 Considering the rate of recidivism and tokeep even with this 10 percent rate of in-crease, 30 percent of all alcohol depen-dent persons should be treated in a givenyear.

4 This figure should be divided into twobecause alcohol dependent personsconstitute only half of the in-need popu-lation. Thus, 15 percent of the overallin-need population should be treated ina year.

5 Add a 5 percent buffer to do more thankeep pace with the growth of the prob-lem. Therefore, 20 percent of the prob-lem drinkers per year are considered asthe target population.

One of the most serious problems with pre-scriptive models is that the assignment ofassumptive values to the estimated popu-lation “in-need” of services is a rather arbi-trary procedure based on empirical datawhich are questionable in terms of reliabil-ity and validity. For example, rates of re-cidivism are estimated from treatment data.The figure of 10% to represent the increasein the number of alcohol dependent personsfrom one year to the next is not likely to bea constant. These proportions can be con-sidered at best as very rough guesses. More-over, the values are likely to vary acrossdifferent planning regions and over time.Another problem with the prescriptivemodel is that it can be value laden, espe-cially in those aspects of the model wherelittle empirical data exist to guide the se-lection of various parameters. For example,treatment practitioners will have differentopinions concerning how the demand popu-lation should be apportioned throughout thetreatment syste.


WHO/MSD/MSB 00.2d

Efforts have been made to minimise thissubjective component. A comprehensiveforecasting model for estimating the ca-pacity of alcohol treatment services inOntario, Canada (Rush, 1990) basesthese estimates on six different sourcesof information: published research litera-ture on patient characteristics; cost-effec-tiveness of treatment, and rates of comple-tion of treatment; a preliminary clientmonitoring system for assessment and re-ferral services; a detoxification reporting sys-tem; a triennial provincial survey of alcoholand drug programmes; informed opinionfrom clinical and research experts and anAmerican forecasting model.

Another significant problem with this prescrip-tive approach is similar to that identified forthe more basic continuum of care approach.Specifically, the model will project needs onlyfor services identified a priori as being keycomponents of the ideal treatment system. Thisapproach may restrict innovation in the plan-ning and delivery of services for PSU disor-ders if an outdated, or otherwise inappropri-ately structured, treatment system is used asthe foundation for model development.


WHO/MSD/MSB 00.2d

Workbook 4 provides information aboutprocess evaluation of treatment servicesand systems for PSU disorders. It includesa brief discussion of the evaluation of sys-tem co-ordination. The issues to be ad-dressed, and the measures of co-ordina-tion that may be used, are similar forprocess evaluation and community needassessment. System co-ordination is typi-cally assessed using reports and ratingsfrom directors or managers of agenciesthat are expected to work together in ser-vice planning and delivery. Ratings aretypically given on:

l mutual awareness - the extent to whichstaff know about each other and theirrespective programmes

l frequency of interaction - how often keystaff meet to discuss work-related issues

l frequency of cross referrals - how oftenor how many clients are referred to andfrom different services in the network

l information exchange - the extent towhich services exchange information

Question 4Are existing servicesco-ordinated and what isneeded to improve theoverall level of systemfunctioning?

l staff sharing or exchange - staff of dif-ferent services are permanently or tem-porarily shared or loaned

l other resource exchanges - the extentto which services share funds, meetingrooms, materials or other resources

l consultations and case conferences -exchanges that concern the treatmentof specific clients

l overlapping boards - the number ofmembers in common to communityboards of different services

l normalisation of agreements - the extentto which services have developed formalagreements to co-ordinate activities

Specific measures of service co-ordinationthat may be used in a community needsassessment are not well-developed interms of reliability and validity. One oftentakes a more qualitative approach basedon key informant or focus group interviews.Such qualitative data collection proceduresare described in Workbook 1.


WHO/MSD/MSB 00.2d

It�s your turnPut the information from this workbookto use for your own setting. Completethese exercises below. Remember to usethe information from Workbooks 1 and 2

to help you complete a full evaluationplan. Review that information now, if youhave not already done so.

Exercise 1Think about your treatment programme.List five general areas in which you wantto know more about the needs of the com-munity.

Exercise 2Assess the availability of existing recordsfor each of the areas that you listed above.

Do you have access to:

l morbidity data

l mortality data

Example: What types of services areneeded for cocaine users in the community?1)2)3)4)5)

l number of patients receiving treatmentwithin a certain area and/or treatmentsystem

Your answers to these questions will helpyou to choose needs assessment thatmaximise use of existing data.

Exercise 3Using the information provided in thisworkbook, make the following decisions:

l Decide what method you will useto collect the data (e.g., generalpopulation survey, mortality-basedprevalence model). Review the infor-mation in this workbook as neededto help you decide.

l Choose a sampling procedure forchoosing specific clients/data tosurvey

l Decide the timing of the evaluation

l Develop a procedure for ensuring con-fidentiality and promoting honesty

l Decide who will help you collect data

Exercise 4You will need to prepare an introduc-tory letter and consent form that explainsthe purpose of your study. Review Sec-tion 1A of Workbook 2, entitled, Man-

age Ethical Issues, for more informa-tion about the important topic of par-ticipants rights in evaluation research.


WHO/MSD/MSB 00.2d

Exercise 5Run a pilot test of your evaluation mea-surement and procedures on 10-15 sampleparticipants to ensure that everything runssmoothly. Review section 1c of Work-book 2 entitled Conduct a Pilot Test forspecific information about how to do this.In general, pilot tests assess these ques-tions:

l Do the questions provide useful infor-mation?

In general, all participants should be askedpermission ahead of time before being en-rolled in the study. When you do this, yourshould explain the purpose, nature, andtime involved in their participation. Noperson should be forced or coerced toparticipate in the study.

The standard practice is to have each par-ticipant sign a consent form, which:

l describes the purpose and methods ofthe study

l explains what they will need to do ifthey participate

l explains that participation is voluntary

Example (from above):

Introductory Letter:

We are asking your help in understandingthe needs of the community by filling outa 2 page questionnaire about your sub-stance use patterns. The questions will askabout your substance use and any effectsthat it might have on your life. They willtake about 10 minutes to complete. Allinformation that you provide us will re-main strictly private and confidential.

If you agree to participate, please read andsign the consent form (attached) and re-turn it in the stamped envelope with thecompleted questionnaire. Thank you foryour time.Sincerely,Dr. X

Consent Form:

You agree to participate in a survey ofsubstance use patterns. You will completea 2 page questionnaire, which will takeabout 10 minutes to complete. Your par-ticipation is completely voluntary. You canrefuse to answer any questions and/orwithdraw from the study at any time with-out a problem to you. All your responseswill remain strictly confidential: your namewill not appear on your questionnaire andyour responses will not be linked to youridentity at any time.

I have read the information above andagree to participate.Signature:Date:

Now it’s your turn. Using the exampleabove, and the information provided inWorkbook 2, section 1A, write your ownintroductory letter and consent form.

l Can the questions be administered prop-erly? For example, is it too long or toocomplicated to be filled out properly?

l Can the information be easily managedby people responsible for tallying thedata?

l Does other information need to be col-lected?


WHO/MSD/MSB 00.2d

In this workbook, a wide range of meth-ods have been described that address fourquestions that are commonly asked in aneeds assessment concerning PSU disor-ders. These questions were:

l How many people in the region or com-munity need treatment for PSU disor-ders?

l What is the relative need for treatmentservices across different regions orcommunities?

l What types of services are needed andwhat is the necessary capacity?

l Are existing services co-ordinated andwhat is needed to improve the overalllevel of system functioning?

For each type of question, there arechoices to be made in selecting the spe-cific need assessment models or methods.You must take into account the nature ofthe decisions to be made with the result-ing information and the time, expertise,and resources available. Each model ormethod also has advantages and limita-tions that must be carefully considered.

After completing your evaluation, youwant to ensure that your results are put topractical use. One way is to report yourresults in written form (described in Work-book 2, Step 4). It is equally important,

Conclusion anda practicalrecommendation

however, to explore what the results meanfor your programme. Do changes need tohappen? If so, what is the best way to ac-complish this?

Return to the expected user(s) of the evalu-ation with specific recommendations basedon your results. List your recommenda-tions, link them logically to your results,and suggest a period for implementationof changes. The examples below illustratethis technique.

Based on the finding that over 1/4 of ran-dom sample community respondentshad used cocaine in the past 90 days,and among those, 58% were interestedin receiving treatment, we recommendthat the programme institute a new co-caine treatment service. The serviceshould begin in March, which is tradi-tionally a low-census month for theprogramme, and would allow for extrastart-up time.

Remember, needs assessments are a criti-cal first step to better understanding thePSU treatment requirements of the com-munity. It isimportant to use the informa-tion that needs assessments provide to re-direct treatment services. Through carefulexamination of your results, you can de-velop helpful recommendations for yourprogramme. In this way, you can take im-portant steps to create a ‘healthy culturefor evaluation’ within your organisation.


WHO/MSD/MSB 00.2d

References

hol, Health and Research World, 1989,13(2):178-182.

Robins, L.N., Wing, J., Wittchen, H-U. &Helzer, J.E. The Composite InternationalDiagnostic Interview: an epidemiologicinstrument suitable for use in conjunctionwith different diagnostic systems and indifferent cultures. Archives of GeneralPsychiatry, 1988, 45:1069-1077.

Rush, B.R. Systems approach to estimat-ing the required capacity of alcohol treat-ment services. British Journal of Addic-tions, 1990, 85(1):49-59.

Ryan, K. Assessment of need for alcohol-ism treatment services: Planning proce-dures. Alcohol Health & Research World,1984/1985, 9(2):37-44.

Tweed, D.L., & Ciarlo, J.A. Social-indi-cator models for indirectly assessingmental health service needs. Epidemio-logic and statistical properties. Evalua-tion and Program Planning, 1992,15(2):165-179.

Tweed, D.L., Ciarlo, J.A., Kirkpatrick,L.E., & Shern, D.L.. Empirical validityof indirect mental health needs-assess-ment models in Colorado. Evaluation andProgram Planning, 1992, 15(2):181-194.

Wittchen, H.U., Robins, L.N., Cottler,L.B., Sartorius, N., Burke, J.D., Regiers,D.A. and Participants in the MulticentreWHO/ADAMHA Field Trials. Cross-cul-tural feasibility, reliability, and sourcesof variance in the Composite InternationalDiagnostic Interview (CIDI). British Jour-nal of Psychiatry, 1991, 159: 645-65.

World Health Organization (WHO). Com-posite International Diagnostic Interview(CIDI). Version 1.0. Geneva, Switzerland:World Health Organization, 1990.

Adrian, M. Mapping the severity of alco-hol and drug problems in Ontario. Cana-dian Journal of Public Health, 1983, 74,(Sept-Oct):335-342.

Beshai, N. Assessing needs of alcohol-re-lated services: A social indicators ap-proach. American Journal of Drug and Al-cohol Abuse, 1984,10(3):417-427.

Cottler, L.B., Robins, L.N., Grant, B.F.,Blaine, J., Towle, L.H., Wittchen, H-U.,Sartorius, N., and Participants in theMulticentre WHO/ADAMHA Field Tri-als. The CID-core substance abuse anddependence questions: cross-cultural andnosological issues. British Journal of Psy-chiatry, 1991, 159:653-658.

Cox, G.B., Carmichael, S.J., & Dightman,C.R. The optimal treatment approach toneeds assessment. Evaluation and ProgramPlanning, 1979, 2:269-275.

DeVillaer, M. Client-centred communityneeds assessment. Evaluation and ProgramPlanning, 1990, 13:211-219.

DeVillaer, M. Establishing and using acommunity inter-agency monitoring sys-tem to develop addictions treatment pro-grams. Addiction, 1996, 91(5):701-710.

DeWit, D., & Rush, B.R. Assessing theneed for substance abuse services: A criti-cal review of needs assessment models.Evaluation and Program Planning,1996,19(1):41-64.

Ford, W.E. Alcoholism and drug abuseservices forecasting models: A compara-tive discussion. The International Journalof the Addictions, 1985, 20(2):233-252.

Harris, J.R. & Colliver, J.D. Highlights fromthe 1987 National Drug and AlcoholismTreatment Unit Survey (NDATUS). Alco-


WHO/MSD/MSB 00.2d

The following case examples describe dif-ferent types of needs assessments. As notedearlier in the workbook, most experts agreethat a single, all-purpose needs assessmenttechnique does not exist. This is becauseevaluation planners have different goals, andhave different data resources available.

The first case example describes an evalua-tion of treatments for PSU dependence inBarcelona, Spain. Several computerised da-tabases were already available, and were usedby evaluators to estimate PSU prevalenceand treatment needs within Barcelona. In thisrespect, this case is an excellent example ofhow existing data can be used effectively toconduct needs assessments. The overallevaluation is complex, and includes aspectsof needs assessment, cost analysis (Work-book 5), and outcome evaluation (Workbook7). The planners wanted to know trends inpsychoactive substance use, characteristicsof PSU users, costs of PSU treatment, andeffectiveness of care. Other evaluators in-terested solely in needs assessment could usesimilar techniques in a narrower scope. Touse this technique, computerised data mustbe available.

The second case presents a needs assess-ment that was conducted without theavailability of computerised data re-sources. In this situation, evaluatorswanted to know the service needs for arural and underdeveloped area of SouthAfrica. Official data were unavailable, soevaluators decided to use key informantsurveys and focus groups as their primarymode of data collection. Through meet-ing and interviewing representatives fromgovernment, police, commerce, and thegeneral community, evaluators were ableto determine perceived PSU trends andtreatment needs.

Of note, neither case relied upon clientopinions to assess needs. Direct interview-ing of PS users is another option for needsassessments, and can generate highly use-ful data. Of course, this type of data wouldbe qualitatively distinct from computeriseddatabases and community key informantsurveys. There is no single right or wrongway to assess needs; each technique pro-vides a unique and potentially useful typeof data.

Comments aboutcase examples


WHO/MSD/MSB 00.2d

byRodríguez-Martos, A*Solanes, P.*Torralba, Ll*Brugal, M.T.**

Contact address:Alicia Rodríguez-MartosPlan d’Acción sobre Drogues de BarcelonaPl. Lesseps, 108023-Barcelona (Spain)

Case example of aneeds assessment

This report is based on the research doneon effectiveness of care programmes fordrug dependents by the Barcelona munici-pal drug action plan (Pla d’Acción sobreDrogues de Barcelona) during the lastdecade. Thanks to Barcelona’s Informa-tion Service data were available on theutilisation of outpatient facilities, as wellas population morbidity and mortality sta-tistics. The goal in analysing these facilityand population statistics was to develop

guidelines for programme improvement.The availability of a health informationsystem incorporating data on both servicesand population should allow the assess-ment of accessibility, coverage and effec-tiveness of care.

The Care and Follow-up Centres (CFCs)pertaining to the City Council, have of-fered services since 1990, including theold drug-free programmes and substitu-tion programmes with drug administrationsuch as the methadone maintenanceprogramme. Other therapeutical activitieswere offered including main health care,social, educational and support activitiesfor families, as well as legal advice andattention.

Who was askingthe question(s)and why didthey wantthe information?

The authors alone areresponsible for theviews expressed in thiscase example.

Planning and evaluating outpatientcare for drug dependent patientsin Barcelona (Spain)

* Plan d’Acción sobreDrogues de Barcelona.

** Serveid’Epidemiologia.Institut Municipal dela Salut.


WHO/MSD/MSB 00.2d

Even though Spain had developed a valu-able information system on drug depen-dencies, the Sistema Estatal deInformación sobre Toxicomanías (SEIT),the managers of care centres andprogrammes needed complementary localinformation. This included data obtainedfrom patients’ follow-up. Thus, a globalperspective was adopted based on thatproposed by L‘ðnnqvist (1985), whichstarts with the classical analysis of struc-ture, process and treatment results, con-centrates on the assessment of objectives,coverage and effectiveness of care, takinginto account its cost and secondary effects.

The information system on drugs ofBarcelona, the Servei d’Información sobreDrogodependéncies a Barcelona (SIDB),set up in 1988 under the Pla d’Acciónsobre Drogues de Barcelona, is aprogramme devoted to the systematicanalysis of the size and evolution of drugabuse in Barcelona. It is designed to evalu-ate its size and evolution.

The observation of phenomena withstigmatising characteristics makes it diffi-cult to develop direct measurement tech-niques used for other health problems.Therefore, it was necessary to create aninformation system that, using indirect in-dicators, could enable us to understand andmonitor the drug addiction problem inBarcelona. The process of designing andimplanting SIDB indicators began in 1988following the general outline proposed bythe National Drug Plan and the Drug Ad-diction Plan of Catalonia.

Main care objectives of the Pla d’Acciónsobre Drogues de Barcelona were:

a) to improve quality of life and life ex-pectancy of Barcelona’s drug addicts;

b) to offer enough treatment services sothat access could be guaranteed to ev-ery person asking for it;

c) to get the most out of activitiesfavouring patient’s contact with treat-ment resources as well as changes inaddicts life style and risk behaviours.

Objectives of the evaluation study wereto assess the efficacy and effectiveness ofservices offered in order to accomplishthese main care objectives.

The information system we used for mea-suring and evaluating the achievement ofobjectives set up by the Pla d’Acciónsobre Drogues de Barcelona was theSIDB. The objectives of the informationsystem had been already consolidated andvalidated and their widening was thenconsidered. The additional objectiveswere:

a) to identify trends in drug abuse in thecity of Barcelona;

b) to describe the basic characteristics ofidentified addicts;

c) to support the management, evaluationand implementation of programmes.

The number of people detected by thisInformation System included, for 1995,data on people assisted at emergency room(3,519), treatment starts (4,119), overdose(150), and new identified drug users2,495).

The global cost of this information systemwas 156,456 ECU for 1995.

What resources wereneeded to collect andinterpret theinformation?


WHO/MSD/MSB 00.2d

The SIDB is based on three fundamentalindicators:

1) Treatment starts: with informationfrom first interviews (admissions totreatment) in Care and Follow-up Cen-tres (CFCs) devoted specifically to thetreatment of drug addicts;

2) Drug-related emergencies - the infor-mation of which is obtained from theEmergency Services of main urbanhospitals in the different districts;

3) Mortality - from an acute adverse reac-tion to drugs or overdoses, recorded bythe Anatomic Forensic Institute and bythe National Toxicological Institute.

Treatment startsConcerning first interviews, this indica-tor provides us with information aboutCFCs=ð activity carried out by ten citycentres empowered by local authoritiesto care for addicts. These ten centresincluded four CFCs belonging to theCity Council, and six sponsored by otherorganisations. Under treatment start,we mean:

a) first interview made to the centre bythe person requesting its services;

b) new interview requested by a previouspatient after having interrupted treatmentfor a long time and wanting to start treat-ment again at the same centre. A patientis considered to be new when he/shehasn’t been to the centre for at least sixmonths.

This information is gathered by meansof a standardised survey in each first in-terview. CFCs collect all data in a sys-tematic way in order to provide theSIDB as well as the SEIT, with data on

their activity (type of interview, medicalcheck-up, social and psychological fol-low-up and referrals) and characteristicsof the users (demographic and socio-eco-nomical data, toxic habits and healthcharacteristics). The amount of datagathered from this information networkallows the municipal drug action plan togive priority to some activities, to evalu-ate and to control the management ofthose centres belonging to the CityCouncil.

Drug-related hospital emergencies

The data are actively collected at the Emer-gency Services of the main district hospi-tals by the Epidemiology and Drug Ser-vice nursing team. The information comesfrom the assistance reports recorded by theEmergency Services. An emergency is con-sidered to be drug-related when the dis-charge report either states the person isa user of illegal drugs or when the ini-tials IDU or the words drug addictiónappear on the report. In every identifiedcase, a standardised form for data col-lection, including demographic informa-tion and circumstances of the incident,is filled in. The indicator refers to thenumber of episodes dealt with by hospi-tals and to the number of people beingcared for because of this reason.

Mortality related to acute adverse re-action to drugs (overdose)

Information is obtained from therecords of autopsies carried out by theAnatomic Forensic Institute. Data arecollected monthly by the Epidemiologyand Drug Service nursing team. A caseis registered when the forensic surgeonreports that this death was due to over-dose. The report includes the macro-scopic pathology, the circumstances ofdeath and eventual tools or objectsfound at the scene of death as well asany report given by family or friends ofthe deceased person. The toxicologicalfindings are not taken into account.

How were the datacollected andanalysed?


WHO/MSD/MSB 00.2d

Besides this information about the generalpopulation, data from Barcelona’s prisonfiles have been collected since 1993.

All these data were processed in such away that its validity and consistency couldbe assured:

a) the data collection was carried out byspecially trained health professionals;

b) a protocol had been developed defin-ing concepts and criteria for inclusion; andthis was a reference protocol for all peopleworking in the SIDB at any stage of theprocess;

c) there was a validated entry of data intothe computer.

After examining the reliability and inter-nal coherence of data gained from differ-ent recorded episodes using the chosenindicators, it was concluded that therewas a need for an identifying element thatcould be used to link different registerstogether. The chosen element was the firstthree letters of both surnames (from fa-ther and mother), birth date and gender.Afterwards, we were able to use an algo-rithm for maximising the probability ofunequivocal identification and matchingevery individual with episodes protago-nised by himself. Validity confirmationwas thus achieved in 97% of pairings,with sensitivity and specificity both over95%.

Measurement of Activity, Productivityand Cost of Care

For measuring the activity of treatmentcentres, standardised measurement unitswere used for three types of activities:first interview, follow-up visits andmethadone dispensation. The assessmentof patients in current treatment hasproven to be useful, taking into accountthe diversity of drugs involved and the

necessity of assessing social priorities tobe answered.

There has also been an attempt to developa unit of analysis for alternative produc-tivity, based on product analysis and anestimation of time assigned to profession-als for different care activities. An inter-disciplinary group defined the intermedi-ate care products accomplished by themunicipal CFCs, and mean time neededfor every basic product was then calcu-lated.

Coverage evaluation

An estimation of the target population wasneeded for evaluating the programmes’coverage. Otherwise, it would not be pos-sible to ascertain if the programme werereaching only a small proportion of thepopulation in need. The SIDB providedus with the data required to estimate cov-erage using capture/recapture techniques(Domingo-Salvany et. al.).

This kind of information is most useful forestimating the need for already existingservices and for quantifying the volume ofusers in need of other care services. Giventhe chronic and relapsing nature of addic-tions and to evaluate coverage properly, itwas important to differentiate between firsttreatment starts and patients who startedagain after drop-out, both among admis-sions and patients following treatment.

Assessment of Effectiveness

Various indicators, based on scaleswhich match several variables in an ac-cumulative way, had been proposed tomeasure the efficacy of care. One indi-cator could be the percentage of treatedpatients maintaining abstinence at twelvemonths follow-up. Nevertheless, theevaluation of effectiveness needs to beindirect, using such indicators as reten-tion in treatment programmes; improve-


WHO/MSD/MSB 00.2d

ment in delinquency; overdoses andmortality because of acute adverse drugreaction; utilisation of other medical ser-vices; incidence of tuberculosis andAIDS. It has been suggested that peoplemaintaining contact with care servicesuse illegal drugs less, present less delin-quency and are less involved in legalproblems, even when they are not cured(Buning, E., 1994). Thus, we elected tomeasure the quality and effectiveness ofcare resources by measuring the reten-tion rate in treatment programme. Inrecent years, a greater retention inmethadone maintenance programmescompared to that in drug-freeprogrammes had been proven. In ourstudy we have compared the retentioncapability among four different munici-pal CFCs and between both types oftreatment programmes (methadonemaintenance and drug-freeprogrammes).

Another approach to the measuring of ef-fectiveness could be to register the evo-lution of happenings the avoidance ofwhich is one of the treatment goals: pettycrime and legal offences; utilisation ofother care services (emergency room,etc.); AIDS and tuberculosis incidence;and overdose deaths. Actually, the SIDBprovides us with data concerning over-dose deaths, emergency room utilisation,etc., while data on AIDS and tuberculoseincidence are provided by the Epidemio-logical Service. A deviation in expectedtrends concerning those items may be at-tributable to the impact of new policiesand programmes.

To measure delinquency one could col-lect the reporting of criminal actions.Nevertheless this variable is liable toswings related to police policy or publicopinion and attitude. To avoid this prob-lem, surveys of representative samples ofpopulation have been and are still beingcarried out. The survey on victimisationand urban security is held on a periodical

basis. It provides information oncitizens=ð worries by two different ways:by an open question, with spontaneousanswer, and by a list of topics proposedto the surveyed so that he/she could ratehis/her level of worry related to each one.The follow-up and comparison of resultsamong successive surveys (SabatJð et al.,1997) show interesting changes over theconsidered period of time, probably re-flecting both the dimension of crime andpeople’s perception of drug problems.

According to the last surveys, there is agrowing understanding on the part of thepopulation concerning drug addicts andthe usefulness of rehabilitation policies.In the spontaneous statement about wor-ries, the drug problem has been droppingsteadily (from 10% to approximately 2%).Within the list of topics which might because for concern, drugs have also de-clined, even for people who continue toassociate youth violence with drug use,when specifically confronted to this ques-tion. There is a trend towards the reduc-tion of delinquency registered by the po-lice parallel to the starting of methadoneprogrammes (unpublished data).

Measurement of activity, productivityand cost of care

Table I shows defined intermediate prod-ucts for classifying the activity of four mu-nicipal CFCs with concerted manage-ment, mean time estimated in each caseand equivalence in care units for drug de-pendencies.

Table II shows care products and costsstarting from an estimation of direct costsof treatment, without including either in-

What did they find out?


WHO/MSD/MSB 00.2d

vestments or indirect costs. Comparisonof different products among centresshowed the different way they operated.It also indicated that cost analysis centredin time assigned to care activities as wellas that centred on drug dependence careunits (DCUs), dramatically reduced thevariability in estimation of unitary costs.

The mean annual cost per client was 365EDU, within a large range, from 343 to445 ECU among centres, reflecting differ-ent treatment models and the balance oftherapeutic activities. The cost of a firstinterview lay between 49 ECU and 56ECU. Methadone dispensation had a costof 2.59 ECU.

Those centres with a lot of patients in lowretention programmes had more clients atthe expense of a bigger volume of not veryactive users, generating less mean costs,than those attending centres with morepatients in high retention programmes.

Coverage evaluation

By means of capture/recapture tech-niques applied over six months, both totreatment admissions and to emergencyvisits and jail entrances, a prevalence ofaround 10,000 active opiate addictscould be calculated for 1993. From thisinformation, we could estimate the per-centage of those covered by treatmentservices. Based on the number of pa-tients in treatment in our four CFCs, wewere also able to estimate the volumeof drug addicts attending treatmentprogrammes in the whole city, thus in-cluding the remaining six CFCs ofBarcelona. Concerning opiate addicts,5,446 addicts living in Barcelona hap-pen to be in treatment, which representsa 54% coverage of the target popula-tion (4,209 patients in municipal CFCsand 1,237 attending other CFCs).

Table 1: Intermediate products defined for classifying the activity of four CFCs belongingto the City Council with concerted management. Estimated mean time and proposedequivalence in drug dependence care units.

Meantime(in minutes)

Estimatedactivityminutes

Annual activity

Equivalencein drug

dependencecare units

(DCUs)

First visit

Therapeutical (a) follow-up

Medical follow-up visit

Social follow-up

Therapeutical group for families

Therapeutical group for patients

Nurse interview

Drug dispensation

Product

2,015

15,258

18,442

7,925

167

549

3,507

206,295

50

25

15

30

45

45

7

3

100,750

381,450

276,630

237,750

7,515

24,705

24,549

618,885

1.00

0.40

0.40

0.40

0.40

0.40

0.10

0.05

a) Therapeutical refers to any care offered within the recovery programme (mainly counselling and psychotherapeuticalapproach)

CFC: care and follow-up centre; DCU: drug dependence care unit


WHO/MSD/MSB 00.2d

Centre B Centre CCentre A Totalactivity

Product

684

6,053

7,057

1,493

713

65

312

57,617

76,460

1,342

56,975

513,037

149

9,394

8,139

First interview

Sucessive therapeutical (a)follow-up interview

Sucessive medicalfollow-up interview

Nurse interview

Social follow-up

Therapeutical group for families

Therapeutical group for patients

Drug dispensation Costs

Annual Cost (thousands of pts.)

Active users

Cost per user

Estimated minutes of care activity

Cost per care minute (pts)

Drug care units (DCUs)

Cost per DCU (pts.)

631

3,020

3,692

1,129

4,376

37

90

55,986

74,259

1,314

56,514

475,286

156

8,029

9,249

353

2,429

2,708

274

1,422

38

137

49,627

48,019

813

59,064

320,329

150

5,555

8,644

2,015

15,258

18,442

3,507

7,925

167

549

206,295

252,943

4,209

60,096

1,672,234

151

29,617

8,541

Centre D

347

3,756

4,985

611

1,414

27

10

43,065

54,205

740

73,250

363,582

149

6,638

8,166

Table 2: Activity and costs of the CFCs belonging to the City Council with concertedmanagement. Barcelona 1994.

a) Therapeutical refers to any care offered within the recovery programme (mainly counselling and psychotherapeuticalapproach)

CFC: care and follow-up centre; DCU: drug dependence care unit

Table III (on page 36) presents this infor-mation related to municipal CFSs between1991-1994. The number of people enter-ing treatment for the first time went downduring this period to about 25%, whatcould be attributed to a growing numberof drug users getting in contact with thesystem.

Assessment of effectiveness

Clear differences could be observed amongcentres concerning their retention rate: af-ter 2 year follow-up, retention was 77%for patients on methadone and 6% for

those on drug-free programmes. Theseresults should imply a substantial redefini-tion of goals and objectives for the cen-tres, as well as a review of inclusion crite-ria for methadone programmes.

With respect to the general population,there has been a reduction in the percent-age of people identifying drugs as one ofthe most important social problems (from9.7% in 1991 to 2.7% in 1993). Coincid-ing with a stabilisation in victimisation, citi-zens tended to consider drug addicts aspatients, demanding more treatment andcare resources.


WHO/MSD/MSB 00.2d

The utilisation of hospital emergency roomsdepends on several factors, including thekind of answer given by the patient. A ser-vice prone to administer or prescribe cer-tain drugs will automatically increase its in-flow of drug users. Regardless of theattraction exerted by each centre and theannual oscillations, Barce- lona’s hospitalemergencies have reduced to around 20%between 1988 and 1993 (Table IV).

Tuberculosis and AIDS are monitored inthe surveillance system, both diseases be-ing strongly related. After an increase from1988, tuberculosis and AIDS had both de-creased. Tuberculosis in intravenous drugusers (IDUs) increased 47% between 1988and 1992 (from 155 to 228), descendingagain the year after. Prevalence of tubercu-losis remained stable, showing a decrease

in non-IDU population. In 1993, 177 newcases of tuberculosis were declared in IDUs(see Table V). Nevertheless, 314 IDU pa-tients with tuberculosis were registered inBarcelona concerning chemotherapy admin-istered to them during the year (part of themwere patients notified the year before andcurrently following treatment; others werepatients who had dropped out of treatmentand were lost for follow-up). Informationon tuberculosis in different populationgroups was gathered since 1987.

The spread of HIV infection among IDUshas partly been responsible for the increasein tuberculosis rates. Another consequenceof this infection is obviously the rise inAIDS cases among IDUs declared inBarcelona residents. Between 1988 and1993, while the definition of case by the

%Startingtreatment forthe first time

Totalstarts

1991

1992

1993

1994

Product

1,099

1,243

735

695

44.9

47.2

29.6

25.4

2,448

2,633

2,483

2,736

Table 3: Total treatment starts in four CFCs* belonging to the City Council. Barcelona,1991-1994.

CFC: care and follow-up centre

Table 4: Illegal drug-related emergencies attended by four university hospitals withpermanent emergency ward. Barcelona, 1990-1993.

MarVall Hebron Totalstarts

Year

1,099

1,243

735

695

2,318

2,010

1,832

1,541

5,065

5,078

4,520

3,823

1990

1991

1992

1993

Clinic

1,099

1,243

735

695

Sant Paul

1,099

1,243

735

695


WHO/MSD/MSB 00.2d

Centres for Disease Control (CDC) wasstill in force, AIDS cases soared. In 1993,229 cases were declared in the city, 47%more than in 1988. In 1994, there was atop incidence coinciding with the change

in AIDS definition, which meant the in-clusion of new TBC cases as AIDS. Af-terwards, there was a drop in incidencewith a trend to stabilisation around levelsof 1990 (Table VI).

RatesCasesYear

854

1,042

923

1,016

1,129

1,101

999

979

899

8,942

50.2

61.2

54.2

59.7

68.7

66.9

60.8

59.6

54.7

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

Cases

87

155

161

213

216

230

177

228

165

1,632

Rates

5.1

9.1

9.5

12.5

13.1

13.9

10.8

13.9

10.0

Tuberculosis in IDUs Total Tuberculosis

Table 5: Tuberculosis incidence in IDUs (a) and main population. Barcelona, 1987-1995.

Rates for 100,00 inhabitants; (a) IDU: intravenous drug user

RatesTotal AIDScases

AIDS % inIDUs (a)

Year

267

358

434

452

505

460

667

558

3,918

15.7

21.0

25.5

27.5

30.7

27.9

40.6

33.9

58.4

53.6

55.1

47.6

49.3

49.8

57.7

51.4

52.3

1988

1989

1990

1991

1992

1993

1994

1995

Total

AIDS causesin IDUs

156

192

239

215

249

229

385

287

2,051

Rates

9.1

11.3

14.0

13.1

15.1

13.9

23.4

17.5

Table 6: Annual evolution of AIDS cases in drug addicts and of total AIDS cases.Barcelona 1988 to 1995. Data by 30.06.1996

IDU: Intravenus drug user; a) % related to the yearly total AIDS cases.


WHO/MSD/MSB 00.2d

AIDS cases reflect infections receivedseveral years before. Therefore, it seemedbetter to analyse infections among caredpatients. Recent estimations on HIV in-fection rates among drug dependents incontact with Barcelona’s treatment cen-tres, provided an incidence of 4.8 in-fection/100 people/year of follow-up.There has been a trend towards reduc-tion: from an incidence rate of 6.24 in1991 to a rate of 3.46 in 1995. Theseare big figures, but similar to thosegiven by the USA in IDUs (4 people ayear). Compared to rates calculated atan European level, ours are lower thanthose of Italy (7.4 among IDUs in treat-ment and lower than the annual HIV in-fection incidence rate (11.7) estimated

for IDUs in contact with AIDS preven-tion programmes in Alicante (Spain) be-tween 1987 and 1992. However, thereis a need for critical appraisal whencomparing data; indeed, several datasources suggest that, in every popula-tion, frequency of HIV infection goesdown after a period of high incidence,even without preventive interventions.

Deaths because of overdoses increasedbetween 1988-1994 and tended to de-crease afterwards. Compared to mortal-ity in other cities, Barcelona presented ahigher frequency of overdose deaths; onepossible explanation being our higherprevalence in intravenous administration(see Figure 1).

Figure 1: Three-monthly evolution of mortality due to acute drugadverse reaction. Total number and mobile mean 4th trimester 1994.

Source: Institut Anatomic Forense de Barcelona(*) Mobile mean: arithmetical average between number of deaths in the previous and followingtrimester

*


WHO/MSD/MSB 00.2d

How were the resultsused?

After analysing results, it was concludedthat several changes had to be introduced.For example, it was seen as necessary topotentiate methadone programmes (num-ber and availability: low thresholdprogrammes, methadone bus, as well asevery resource devoted to harm reduction(syringe exchanges, etc.) There was a needfor potentiating medical care (vaccination,chemoprophylaxis, early treatment and fol-low-up) and social awareness had to be pro-moted so that patients and programmes (es-pecially harm-reduction approaches) couldbe better accepted. Since 1995,programmes have been launched tailoredfor each urban district to answer their ownand differentiated needs. The Pla d’Acciónsobre Drogues de Barcelona cares for thefurther development of this territorialproject promoting the direct involvementof district authorities and neighbouring as-sociations so that everybody is able to feela personal participation and to make everystep together.

A further change to be introduced was themodel of contract, this time according tothe delivered care services. This wasplaced under management of the Pland’Acción sobre Drogues de Barcelona. Aprotocol was established devoted to thefollow-up of the contract. This considersthe different types of treatmentprogrammes and the minimal capacity for

patients to be cared. The various compa-nies and NGOs based in Barcelona whichare capable of offering treatmentprogrammes for drug dependencies mayopt for our contract by means of present-ing their technical project to a public com-petition. This project has to be in agree-ment with the protocols established foreach CFCs in the bases of the competition(specifications). Each treatment centre inthe city has to take care of patients be-longing to its area of influence, which hasto be previously designated. According tothe analysis made by means of the SIDBin those areas, each centre gets assignedits priorities on types of programmes andon the number of patients to be cared. Theunit for the calculation of the budget inevery single centre is established throughthe drug dependence care units (DCUs).This is a unit of productivity, calculatedthrough the assignment of times for everytherapeutical intervention according to theallotted programme and to the technicalprotocol which has been agreed upon(number of visits and recommended typol-ogy). On the other hand, the technicalspecifications lay down some quantity andquality standards (retention, coverage,etc.). Companies overcoming these stan-dards may apply for a reward in the formof a greater payment every three months.

Only the high degree of development ofthe SIDB information system has allowedthe working out of these technical con-tracts and their adjustment to the prob-lems of each area of the city.


WHO/MSD/MSB 00.2d

It�s your turnWhat are the strengths and the weaknesses of the presented case example? List threepositive aspect and three negative aspects:

Strengths of the case study

1

2

3

Weaknesses of the case study

1

2

3


WHO/MSD/MSB 00.2d

References for case example

Aviño-Rico, MJ., Hernández-Aguado, I.,Pérez-Hoyos, S., García de la Hera, M.,Bolúmar-Montrull, F. Incidencia de lainfección por VIH-1 en usuarios de drogaspor vía parenteral. Med. Clín.(Barc.),1994, 102: 369-73.

Brugal, MT., Torralba, L., Ricart, A.,Queralt, A., Graugés, D., Caylà, JA.Evaluación de los programas detratamiento por toxicomanías. Contri-bución del análisis de supervivencia. Gac.Sanit., 1994, 44 (suppl.): 30.

Brugal, MT., Caylà, JA., García de Olalla,P., Jansá, JM. Disminuye la infección por elvirus de la inmunodeficiencia humana enlos drogadictos intravenosos de Barcelona?Med. ClRðn. (Barc.),1995, 405: 234.

Brugal, MT., VillalbRo,Torralba, L.,Valverde, JL., Tortosa, MT.Epidemiologia de la reacción agudaadversa a drogas. Barcelona, 1983-92:análisis de la mortalidad. Med. Clon.(Barc.), 1995, 105: 441-445.

Brugal, MT., Graugés, D., Queralt, A.,Ricart, I., Caylà, J.A. Sistemad’Informatión de Drogodependències deBarcelona (SIDB). Informe 1994. Serveid’Epidemologia. Institut Municipal de laSalut. Pla Municipal d’Acción sobreDrogodependències. barcelona, 1996.

Domingo-Salvany, A.; Hartnoll, R.L.;Maguire, A.; Brugal, MT; Albertin, P. Andthe prevalence study group (Caylà, J.A.;Casabona, J.): Analytical considerationswith capture-recapture prevalence estima-tion: case studies of estimating opiate usein Barcelona metropolitan area. Ameri-can Journal of Epidemiology, (in press).

Institut Municipal de Salut. Serveid’Epidemiologia. Plan Municipal d’Acciónsobre Drogodependències. EvolucióIndicadors SIDB. Sistema d’Informaciósobre Drogodependències a Barcelona, 2ontrimestre 1996. Ajuntament de Barcelona.Área de Salut Pública, Juliol 1996.

Lönnqvist, J. Evaluation of psychosocialtreatments. Acta Psychiatrica Scandinavica,1985, 71 (Suppl. 319): 141-150.

Manzanera, R., VillalbRð, JR., Torralba,L., Solanes, P. Planificación y evaluaciónde la atención ambulatoria a lasdrogodependencias. Med. ClRðn. (Barc.),1996, 107: 135-142.

Nicolosi, A., Musicco, M., Saracco, A.,Molinari, S., Zikiani, N.: Lazzarin, A. In-cidence and riskfactors of HIV-infection:a prospective study of seronegative drugusers from Milan and Northern Italy,1987-89. Epidemiology,1990,1: 457-459.

Sabaté, J.; Aragay, J.M., Torrelles, E. Ladelinqüència a Barcelona: realitat I por.Catorze anys d’enquestes de victimitzacio(1984-1997). Barcelona, Inst. Estudis Met-ropolitans de Barcelona. Area Metropolitanade Barcelona. Mancomu-nitat de Municipis.Ajuntament de Barcelona,1997.

Sánchez-Carbonell, J.: Camí, J.Recuperación de heroinómanos: defini-ción, criterios y problemas de los estudiosde evaluación y seguimiento. Med. Clín.(Barc.), 1986, 87: 377-382.

Vlahov, D. The ALIVE study. HIVseroconversion and progression to AIDSamong intravenous drug users in Balti-more. In: Nicolosi, A. (de). HIV epidemi-ology. Models and methods. New York:Raven Press, 1994, 31-50.


WHO/MSD/MSB 00.2d

Case example of aneeds assessment

A study to determine the welfareservice needs in the EasternTransvaal, Republic of South Africa

ByM. K. ChristianDirector: Professional ServicesNational Deputy Executive DirectorSANCA National

The authors aloneare responsible forthe views expressedin this case example.

Who was asking thequestion(s) and whatdid they want to know?

The Eastern Transvaal region (now namedMpumalanga - one of the nine Provincesof the Republic of South Africa) is a verybig and largely underdeveloped area. Out-side of a few developed urban and indus-trial areas, there is a farming communityand a tourist industry as this Province in-cludes the famous Kruger National Parkand a number of other scenic areas. So-cial Welfare Services and facilities werealmost non-existent for the majority of theblack population.

The responsibility for the area concernedfell under the Regional Welfare BoardEastern-Transvaal who according to theNational Welfare Act (Act 100 of 1978)had to:

a investigate the social problems which oc-cur in its region and consider, plan andpropose measures for the solution thereof;

b determine of its own accord or on requestthe existing or future welfare needs of theinhabitants of the region or any partthereof;

c plan and prepare a welfare programmewith a view to future development orprovision of welfare services/facilitieswhich would be likely to be necessaryto satisfy such

(i) identified needs,(ii) and to recommend the order of priority in which such services should be accorded;

d up to 1990, the local government, theTransvaal Provincial Administration(TPA) was the main role-player render-ing only social welfare services at grass


WHO/MSD/MSB 00.2d

4 act as a link between communities andspecialist services;

5 co-ordinate welfare services locally.

In order to carry out the above mission,the TPA Social Work Services decided todraw in the NGO’s initiative in the follow-ing manner:

1 Thirty one (31) additional social work postswould be made available. These socialworkers would be appointed by the TPAand would be employed by them;

2 The NGOs would then be allocated acertain number of posts and would beinvolved in the recruiting, training, su-pervision and evaluation of the socialworkers concerned;

3 It was envisioned that after 2 to 3 years,the social worker posts would be takenover by the agency concerned.

The Regional Welfare Board of the East-ern Transvaal initiated the need assessmentinvolving the TPA and the SA NationalCouncil on Alcoholism and Drug Depen-dence (SANCA). This was conductedamong the communities of the rural andunderdeveloped areas of the EasternTransvaal in order to assess, the role ofalcohol and drug use within the broaderissue of social problems identified. This in-formation was not to be used in isolationof other problems but to become an inte-gral part the planning of a 5 year welfareprogramme for the communities con-cerned. It was anticipated that theprogramme would move patiently throughthe following three phases:

Phase 1 - social planning phase: Com-munity profile and needs assessment;

Phase 2 - the community developmentphase;

Phase 3 - the service developmentphase.

roots level. Specific Services in thefields of:

1) physical disability: blind, deaf,cripple;2) care for the aged;3) mental illness/health;4) substance abuse;5) child and family welfare;6) offender rehabilitation

were almost non-existent because thesewere usually rendered by the Non-Govern-ment Organizations (NGOs/registered wel-fare agencies), who had neither the moneyor subsidised social work posts to carry outsuch services in the small widely separatedcommunities, over such a large area.

In 1991, the Regional Welfare Board andthe TPA invited the leading NationalCouncils NGOs specialist-agencies pro-viding specialist services, to participatein a think-tank and workshop, as to howto move away from grass roots socialwork, towards enabling and assistingNGO/specialist agencies to develop com-munity structures and render muchneeded services. Following this historicmeeting, the outline for a unique 5 yearsocial welfare development programmewas established. In this programme, vari-ous specialist NGO agencies would be as-sisted by the Regional Welfare Board andTPA to:

1 investigate social problems and determinelocal welfare needs. Listen to the com-munity members and involve them in wel-fare activities;

2 act as facilitators to bring people or groupsof people together, to address local wel-fare problems;

3 inform communities regarding welfarepolicy and other matters (e.g., registra-tion and subsidisation of child welfarefacilities);


WHO/MSD/MSB 00.2d

The Community profile and needs assess-ment was to be given priority and carried outduring the normal hours of employment andwhile social workers were visiting communi-ties concerned. Therefore, cost were mini-mal and limited to training, where travel andaccommodation were paid by the TPA.

Planning

Phase I: The social work section of theTPA Eastern Transvaal drew up the con-cept document of agreement between theTPA and the agencies - this was to be dis-tributed before July 1991.

1 a concept service contract for the socialworkers was to be drawn up by the TPAregional office and distributed to theagencies concerned before July 1991;

2 all documents were to receive approvaland clearance before commencing theneeds assessment. The remaining postswould be filled at a later stage in theprogramme.

Training: A training programme was de-veloped to provide specialist input fromthe various agencies. The group of 22 so-cial workers met at a venue inJohannesburg for 6 days, additional train-ing in community development was alsogiven.

Except for the specialist input, all socialwelfare staff were employed to become anintegral part of the entire social welfareprogramme, beginning with the commu-nity profile and needs assessment.

Further on in the programme and ac-cording to the needs/priorities identifieda number of social work posts were ac-tually allocated to the agencies - whothen proceeded to provide special train-ing to enable the workers to address theproblems with the community. One so-cial work post per 20 000 people wasdecided upon.

Time frame

Although it was planned that by 1996 so-cial welfare programmes would be estab-lished according to the needs expressed,the phases could not be neatly boxed intime. The Social Planning phase wouldprobably be ongoing while CommunityDevelopment was being initiated. In thesame way, the Service Development phasemay begin during the continuation of theCommunity Development phase.

What resources whererequired?

During the Social Planning phase each of the8 agencies and 2 Government Departmentsparticipating, allocated official representativeswho formed the core group together with theexisting social work staff of the TPA com-munity services - in all some 30 persons.Each group agreed to bear their own costsand provide specialist input. The TPA inWitbank agreed to provide the secretariatand co-ordinate the planning and follow-upmeetings. Twelve social work staff were al-ready doing community work, 10 more wereselected and employed. Care was taken toemploy workers who were familiar with theregion and the various cultural groups andlanguages represented.

1 the social workers appointed were to beemployees of the TPA and would receivetheir salaries from the TPA; the Commu-nity profile and needs assessment wouldbe undertaken as a priority and as part ofofficial duties;

2 the TPA would supply offices and ve-hicles;

3 the specialist agencies would be respon-sible for professional supervision and in-service training of the social workers.


WHO/MSD/MSB 00.2d

How were the datacollected?

The community profile and needs assessmentquestionnaire was drawn up, keeping in mindthat unlike urban areas, official data was un-likely to be available and facts and opinions(quantitative/qualitative) would have to becombined with the questionnaires. All pro-files, and completed questionnaires and re-ports would be in English. Group and indi-vidual interviews would be conducted in thelanguage of choice: Zulu, Swazi, Ndebele,Sotho and Xhosa.

The region was divided up and variouscommunities/townships were assigned tothe 22 social workers. Accessing the com-munity, identifying existing infrastructure and

key community members, professional andlay people who could provide information,kept the workers occupied for severalmonths. Trust had to be built up as well. Acommunity profile and needs assessmentwas to be completed for each community.

In all 15 communities that were surveyedover 3 years, more than 2000 structuredinterviews took place - finalising into 1community profile with a detailed reportand recommendations for each area.

Some information and statistics was ob-tained from visiting Town or Village coun-cils, police stations, churches, clinics,schools and consulting records.

Other information had to be obtained throughinterviews with key people involved such as:

PopulationMagisterialDistrict

20000

6510158994

23286

477449318

4600

170004068

2670

57840130048

8000

27300

24504

Piet Retief

Witbank

Bethal

Standerton

Wakkerstroom

Volksrust

Perdekop

Evander

Sabie

Lydenberg

Barberton

Community/Township

e-Thandukuyklamya

ThubelihleKwaguqua

e-Mzinoni

SakhileThuthukani

e-Sizameleni

VukuzakheMorgenzon

Siyazenzela

Lebolance-Mbalenhle

Simile

Mashishina

c-Mjindini

Actual community Townships surveyed N=15


WHO/MSD/MSB 00.2d

the local shopkeeper, shebeen owner (indig-enous tavern), Induna (minor chieftain).

Community group meetings were held anddiscussions initiated - not only did a valu-able community profile emerge, but factsand opinions were sought on a number ofissues. Group meetings were popular -providing an opportunity for Communityto get together and enjoy refreshments(this was minor but a most important costin the programme).

The questionnaire on alcohol and drug usewas very comprehensive, target groups ofrespondents came from clinics, health careworkers, nurses and doctors and other so-cial workers, traffic departments, police,magistrate courts, teachers, ministers of re-ligion and members of the community andyouth. Sometimes the workers left a ques-tionnaire to be completed - in most casesbecause of language and literacy difficul-ties, these were completed by the socialworkers. Availability and willingness of re-spondents to participate were the only cri-teria used. No resistance was experienced.

The social workers received regular super-vision and encouragement’s. Reports andcompleted work was finally co-ordinatedby the TPA officials had core group.

How were the dataanalysed?

A monitoring group was established andmeetings were held every 6 months, toevaluate progress. Each worker was as-sisted to collate all documents pertainingto the Community profile. Gaps wereidentified for workers to proceed withgaining additional information.

Each worker was responsible for the fi-nal profile and report back on each com-munity. In this regard the value of pre-training and the joint effort made in the

preparation of the standard community pro-file, needs assessment and substance abusequestionnaire proved invaluable in beingable to organise the final reports. Very of-ten it was not possible to get statistics orconcrete facts - only general perceptionsand informed opinions. What was most im-portant was that there was seldom any con-tradictions - opinions were firmly held.

Sophisticated computer analysis was notavailable and in many cases would nothave been meaningful because of the na-ture of the data gathering. Individual com-pleted profiles and reports were analysedas available by the core group and the rec-ommendations of various communitymembers and social worker concernedwere taken into consideration for Phase2 and prior to Phase 3.

What did they find out?

The Community Profile and needs assess-ment was able to pinpoint very specifically:

1 the number of people involved and therequirements concerning the needs ofthe blind, deaf and physically disabledand mentally handicapped. This variedonly according to the size of popula-tion.

2 In general however, no matter the sizeof the population, it became very clearthat all communities share and identi-fied the same social problems and thatthe underlying theme expressed overand over again was that of “unemploy-ment, poverty and alcohol abuse werethe social problems which go hand inhand”.

Substance use/ abuse

There were two main substances used -alcohol and dagga (cannabis satavia). Glueand petrol sniffing were very minor.


WHO/MSD/MSB 00.2d

Alcohol

Commercial products were very expensiveand usually purchased from Bottle Stores ordrunk in Taverns. Because of unemploymentand poverty, cheap alcohol concoctions orhome-made brews called SPION orMBAMBA were available at Shebeens.These are informal backyard type taverns,usually not registered. Shebeens are themost important adult recreational facilityand a source of income for the owner.Youth also obtain their liquor from theshebeens in the afternoons after school, butdo not drink on the premises. Women visitthe shebeens in the late morning and earlyafternoon, while the men dominate theevening and night sessions. It was esti-mated that more than 60% of the popula-tion in all 15 townships were abusing al-cohol and of this number, 40% were drinkingat alcoholism levels.

Problems of alcohol abuse in thecommunity

Alcohol was abused over weekends andafter working hours by youth, men andwomen between the ages of 20 and 50.The confirmed opinion is that more mendrink than women, but that women arenow drinking more than they used to, ahome brew of soured milk or sorghum wasconsidered an important part of the meal.Poverty prevented the woman from pro-ducing this, thus malnutrition was high.

1 wife battering and physical assaults werehigh

2 families kept impoverished through drink-ing and unemployment

3 most women do not recognise that a hus-band may have a problem. They live withinthe circumstance and accept the alcoholuse/abuse as a way of life.

4 hospitals and clinics report high incidentsof assaults.

5 most arrests are for assault, drunk anddisorderly and theft.

6 there were very few cases of drunkendriving, as there was little opportunityand few cars to drive.

7 drunken pedestrians caused accidents -high incidence in rural and country areas.

8 schools report drinking among the highschool groups where alcohol use is asource of entertainment and bought inthe afternoons from the shebeens.

9 little or no drinking or drugging was ob-served in the schools, only 1 case wasreported in all reports.

10teachers however see neglect, malnutritionand signs of physical abuse, they hear aboutconditions at home and they are convincedthat drinking is a serious.

11the opinion most often expressed was thatchildren from these families end up drink-ing themselves and not fit for the labourmarket.

12most epileptics and tuberculosis patientsneglected their medication because of theeffort and distance to hospitals in urbanareas and drank as a form of self medi-cation.

Dagga or Cannabis

The dagga plant is indigenous to SouthAfrica and easily grown for private use.In the communities surveyed, it was notgrown commercially (but is elsewhere inSouth Africa). People were aware of thelegal consequences. Informed commu-nity members and leaders spoke out inunison against legalising dagga as it wasobserved and an importantly held opin-ion that dagga use promoted the “dete-rioration of society”. School childrendropped out and led useless lives orlanded in prison. Prison statistics clearlyindicated that arrests and convictionsdue to dagga were significantly higherthan those of alcohol and usually linkedto crime. It was also noted that the so-phisticated use of dagga and crushedmandrax (white pipe) as smoked in ur-ban areas was virtually unknown in thesecommunities.


WHO/MSD/MSB 00.2d

Dagga is used more by youth as it is cheap(free) and exciting but is often continued intoadulthood. Youth however, did not see daggasmoking as serious, starting fairly early between10 and 20 years with most users between theages of 20 and 30, mainly male.

Generally, people were unaware of servicesor programmes that could help reduce theuse of alcohol and dagga and prevent thesocial and health problems occurring. Whenhealth and social functioning deteriorated, thecommunity managed this within their ranks.In some of the communities, help was avail-able and alcoholics/dagga addicts could bereferred to Themba Centre or dealt withthrough local health clinics where someknowledge was beginning to filter through,TPA social workers throughout the regionhad a case load of less that 30.

Community problems evaluated

The most serious problems identified were:

1 Unemployment

2 Poverty

3 Lack of Infrastructure

4 Alcohol and Dagga Abuse

family and community being negativelyinfluence by these.

In communities, a greater percentage ofthe children were in the care of grand-parents who could not always providecontrol or for their financial needs. Childneglect and abandoned children were afurther indication of poverty as parentsleft to go to the cities to look for work.

The lifting of the influx control legislationa few years earlier had a tremendous effecton the population in these communities,placing tremendous strain on the few exist-ing resources, as people tried to get nearerto work opportunities. By 1994, it was esti-mated that 60% of the working population

was unemployed and that the population hadmore than doubled the 1985 figures.

The political climate in communities andtownships was as uncertain as the politi-cal development in South Africa. Where thetraditional Induna systems were still in op-eration, there was strong willpower toorganise themselves.

Schools were overpopulated and grosslyunder served - influencing future educa-tion and employment opportunities. Pri-mary schools outnumbered high schools6 to 1. In 4 areas there were no highschools. School was also very basic offer-ing no additional skills or training.

Alcohol use was obviously an importantpart of entertainment, used by youth andadult members of the communities. Thisis seen in the extra-ordinary high numberof shebeens (340), taverns (32) andbeerhalls (6) around compared to shops(22) and churches (46). The communityleaders, however, did view the drinking asa serious set back to development and re-quested awareness and educationprogrammes as urgent.

Very limited sports (4) and recreation fa-cilities (2) were found - usually only in themining villages. Cinema and TV almostnon-existent due to weak power supplyand poverty.

Religion played a big role in keeping thecommunity together and was very accept-ing of all conditions of life. Religious lead-ers still had the respect of the communi-ties and even \political bodies and theywere usually the backbone of those in-volved in problem solving.

The needs for adequate shelter, water,roads, electricity were expressed moreurgently that the needs for services forphysically disabled, etc. There was how-ever, an expectation that there should beprovision for these.


WHO/MSD/MSB 00.2d

The youth, not affected by alcohol and daggause, demonstrate a willingness to get involvedwith community issues as well as to organisetheir own entertainment. They appear to beimpatient with older members of the commu-nity who demonstrate apathy to get involvedwith health and welfare issues, especially whenthere is no financial gain. In spite of this com-mon trend, there are community members whodo involve themselves, but require motivationand financial and practical support.

How did they use theinformation?

For the first time in South Africa, the roleof alcohol and dagga in keeping peopleand communities underdeveloped wasdemonstrated, both contributing to causeand effect of poverty, unemployment, etc.Large scale community development wasrequired before the development of spe-cialist services. However, communitywork intervention was urgently requiredand could be implemented. TPA socialworkers already involved in the needs as-sessment were allocated to specialist NGOagencies. SANCA was given 5 posts andone supervisor. These were now givenspecific training in substance abuse, prod-uct knowledge and prevention models andpublic speaking. The Community profilesand their own involvement with the com-munities concerned, already indicated suit-able target groups.

It was felt that the most effective strate-gies to combat alcohol and drug abusewould be:

1 to establish an action committee of con-cerned people. Such a committee wouldbe informed and made aware of thefindings of the community project andwould be motivated to become part ofthe solution.

2 training in helping skills and early identifi-cation of users and people in need of helpwould follow. A core group would be

trained to assist at various levels in thecommunity.

3 awareness campaigns were planned for theyouth between the ages of 12 and 25 tocapture their interest. In a very short while,they formed into SANCA youth groupswhere attention was given to a comprehen-sive life skills programme aimed at their ownstated needs.

The youth quickly cottoned onto thefact that a life of alcohol and drug usewould only continue the misery formany other youth. Strategies now in-cluded training selected youth as peercounsellors who could work among theyoung people themselves, who couldpromote a different lifestyle, give talksand workshops at schools, churches butmore importantly in places whereyoung people congregated.

Further training provided helping skillsand early identification of substanceabuse and a referral system of resourcesavailable elsewhere. Positive mindedyouth were targeted and the peer coun-selling movement had its origin in theEastern Transvaal.

4 Where existing infrastructures such as clin-ics, hospitals were identified, the socialworkers visited to created awareness andoffer a training package suited to theirneeds, or those of their clients, a com-mon example was the pre-natal clinicsvisited by mothers-to-be.

5 Conditions were most often verysimple and lacking any refinements,the social workers ‘ having to gowell prepared to get the messageover to a target group that largelylacked previous formal educationand could not read.

6 Social workers had to be careful not tocreate unrealistic expectations in thecommunity — but to work with what waspossible with maximum utilisation of com-


WHO/MSD/MSB 00.2d

munity members, but at the same time ar-ranging meetings and putting them in touchwith prospective or available resources.

7 The youth to youth movement had openedup many more opportunities to combatalcohol and drugs and had assisted thesocial workers beyond the initial plannedintervention.

a) on-going training and motivationwas required and the developmentof a training curriculum;

b) when funds were available, identi-fying badges, caps and T-shirts wereprovided;

c) the slogan “say YES to life and NOto drugs” was adopted with theSANCA lo o.

At the present time and because of the flex-ible time-frame mentioned previously, allof the above steps have been taken andare being met at various levels in seven ofthe 15 communities.

Given time, the social workers will giveless time to the established programmesand move into the next areas. There is nowa waiting list of sorts, as more and morerequests for similar programs to be estab-lished are being received in respect of Al-cohol and Drug strategies already in op-eration. Other information generated bythe community profile has led to SocialWelfare programmes being developed tomove, into Phase 3 in some of the com-munities concerned.

There was sensitivity to the fact that allthe Community/Townships surveyed wereunderdeveloped, plagued by poverty andunemployment. Care had to be exercised,to avoid labelling the community in anymanner.

In the final analysis, it was the communitymembers themselves who outlined and un-derlined and named their problems. All re-source persons gave their names willingly andonly a few respondents asked not to benamed in person - this was respected. So-cial workers were well received as personswho were trying to help make a difference.No problems regarding the need assessmentwere encountered, only those of distance,long hours and the continual evidence of manyneeds to be met. Monitoring and support forthe social workers were never neglectedthroughout the years.

The Eastern Transvaal Region has nowbecome an official Province of the Repub-lic of South Africa with its own Depart-ments of Health and Welfare. All strate-gies and welfare programmes mentionedhave been adopted and programmes con-tinued in co-operation with the agenciesconcerned.

The model used for the communityprofiles and needs assessment will be car-ried over to other communities inMpumalanga.

1 primary and secondary prevention strate-gies were recognised as a priority;

2 institutional treatment/rehabilitation re-garding substance abuse already existedin the province. Awareness of the needfor treatment and accessibility to the fa-cilities however formed an important partof the strategy:

3 Reconstruction and DevelopmentProgramme (the RDP) of the new Gov-ernment structures has been assistingin the upliftment of these communities,but still has far to go.


WHO/MSD/MSB 00.2d

It�s your turnWhat are the strengths and the weaknesses of the presented case example? List threepositive aspect and three negative aspects:

Strengths of the case study

1

2

3

Weaknesses of the case study

1

2

3


WHO/MSD/MSB 00.2d

References for case example

1 Concept Document: 1991

2 Minutes of Meetings: 1991 - 1994

3 SANCA files on the 15 communities/townships surveyed

4 Community Profiles and Reports: 1992 -1994

5 Progress and Evaluations of WelfareProgrammes initiated as intervention strat-egies (Alcohol and Drugs)

6 Model for Service Development (avail-able on request)

All documents are the property of SA Na-tional Council on Alcoholism and Drug De-pendence but may be utilised for scientificand academic purposes. Any further infor-mation can be obtained on request.

This case example is broader than a typicalneeds assessment. It also includes elementsof process evaluations, cost evaluations, andoutcome evaluations.

Documents

Workbook 3 - apps.who.intthe general steps for evaluation de-scribed in Workbooks 1 and 2. Use this specialised workbook simul-taneously with the foundation workbooks to maximise the