HEPATITIS A ASSOCIATED WITH A HARDWARE STORE WATER … · 2017. 10. 3. · hardware store premises in August and September, 1980. These groups included hardware store employees, family

AMERICAN JOURNAL OP EPIDEMIOLOGYCopyright © 1983 by The Johns Hopkins University School of Hygiene and Public HealthAll rights reserved

Vol. 117, No. 6Printed in U.SA.

HEPATITIS A ASSOCIATED WITH A HARDWARE STORE WATERFOUNTAIN AND A CONTAMINATED WELL IN LANCASTER COUNTY,

PENNSYLVANIA, 1980

G. STEPHEN BOWEN1 AND MARY ANN MCCARTHY2

Bowen, G. S. (CDC, Atlanta, GA 30333) and M. A. McCarthy. Hepatitis A as-sociated with a hardware store water fountain and a contaminated well inLancaster County, Pennsylvania, 1980. Am J Epidemiol 1983;117:695-705.

In August-October 1980, a sudden increase occurred in the number ofcases of jaundice reported among residents of a rural, agricultural section ofLancaster County, Pennsylvania. Investigation confirmed the cases ashepatitis A and showed that the outbreak was associated with consumption ofwater from a water fountain in a hardware store and water from the adjacentfamily home, both of which came from a well contaminated with feces. A totalof 49 cases occurred from August 11 to October 21 in store customers,employees, family members, and persons visiting the family residence. Un-usual features of the outbreak included an epidemic curve with two peaks, avery high attack rate in susceptibles at highest risk (>90%), a high proportionof cases with jaundice (85%), absence of asymptomatic cases, and a highmale to female ratio due to the makeup of the group at risk.

antibodies; hepatitis; hepatitis A; jaundice; water pollution

Common-source outbreaks of hepatitisA have been frequently reported in thepast. Milk, water, food, specific individu-als, institutions, and day-care centershave been identified as sources. Citywater supplies (1), wells, surface watercontamination of local water systems (2),orange juice (3), and Mai Tai (4) have

Received for publication May 19, 1982, and infinal form January 31, 1983.

Abbreviation: HAV IgM Ab, hepatitis A virus IgMantibody.

1 Field Services Division, Epidemiology ProgramOffice, Centers for Disease Control, assigned to thePhiladelphia Department of Public Health,Philadelphia, PA.

2 Division of Epidemiology, Pennsylvania De-partment of Health, Harrisburg, Pennsylvania

Address for reprints: Field Services Division, Epi-demiology Program Office, Centers for Disease Con-trol, Atlanta, GA 30333.

The authors thank the following people for theirassistance during this investigation: Ernest Witte,Robert Gens, Thomas DeMelfi, Robert France, EthelHershey, Mary Fox, John Randle, Doris Yoder,Pauline Zimmerman, Elaine Zuck, Doris Kolb, Mar-ion Jamison, Janet Gantz, Donald Francis, MarkKane, and Charles Schable.

been identified as specific common sourcesof waterborne outbreaks. Most previousreports have relied on clinical and bio-chemical data to establish the diagnosisof viral hepatitis, infectious hepatitis, orhepatitis A.

On September 24, 1980, one of us(MAM), while reviewing surveillancedata, noticed an increase in the number ofinfectious hepatitis case report formsfrom Lancaster County, Pennsylvania,being received at the Pennsylvania De-partment of Health in Harrisburg. Theusual number was one or less per week. Atelephone call to the supervisory publichealth nurse of the local office of thePennsylvania Department of Health inLancaster, which is responsible for Lan-caster County, revealed 13 cases. Six ofthe early cases were in employees of asingle hardware store in northern Lan-caster County. An epidemiologic investi-gation was begun and the data generatedare the basis of this report.

695

Downloaded from https://academic.oup.com/aje/article-abstract/117/6/695/125850/HEPATITIS-A-ASSOCIATED-WITH-A-HARDWARE-STORE-WATERby San Diego State University Library useron 18 October 2017

696 BOWEN AND McCARTHY

MATERIALS AND METHODSDocumentation of hepatitis A or infec-

tious hepatitis cases in Lancaster Countyfrom November 1979 to October 1980 wasdone by reviewing case reports on file inHarrisburg and at the Lancaster office ofthe Pennsylvania Department of Health;by reviewing medical records and dis-charge diagnosis, emergency room, andchemistry laboratory records at EphrataHospital, Lancaster County; by telephon-ing a sample of area primary-care physi-cians and sending others a letter request-ing rapid reporting of cases; by contactinginfection control nurses and infectious dis-ease physicians at area hospitals; and byasking ill persons for names of other simi-larly ill persons.

Initially, all patients were interviewedby telephone or in person to obtain clini-cal data, to discover previously unknowncases, and to obtain information about so-cial gatherings, travel, employment, foodconsumption, sources of food and water,medications, and drug usage. A total of 38patients or adults in patients' familieswere interviewed 4—5 weeks after onsetof illness to detect secondary cases inhouseholds of primary cases. We found178 persons to be household contacts ofthe cases. Questioning was directed to-ward discovery of milder, non-ictericcases of hepatitis.

Initial interviews with all cases showedthe hardware store and adjacent familyhome to be the source of this outbreak (seeResults). Once this was established, weinvestigated the source and mode oftransmission of the outbreak by contact-ing groups of people who had visited thehardware store premises in August andSeptember, 1980. These groups includedhardware store employees, family mem-bers of the owner of the store, members ofa youth choir which met every other weekat the home of the owner of the store, agroup of girl choir members who attendeda slumber party at the family home, andhardware store customers. A list of credit

customers was obtained from the owner ofthe hardware store and every fourth per-son on the list was interviewed if theycould be reached by repeated telephonecalls. No list of cash customers was avail-able.

Blood specimens were collected from illand well store employees, family mem-bers, youth choir members, and 67 percent of known cases 4—8 weeks afteronset of illness. Blood specimens weretested for presence of hepatitis A-specificIgG and IgM antibody by the HepatitisLaboratories Division, Centers for Dis-ease Control, Phoenix, AZ, using com-mercially available radioimmunoassaykits (5) (Abbott Laboratories, N. Chica-go, ID.

Environmental investigation

On September 9, 1980, personnel of theDepartment of Environmental Resources,Pennsylvania Department of Health, ex-amined the well which supplied thehardware store and family home, tookwater samples from the water fountain,and surveyed the septic and drain fieldsystems of the store and family home. OnSeptember 22, water samples were takenfrom the well before the well water enteredthe pressure tank. Samples were alsotaken on September 22 from two wells lo-cated on adjacent properties. Water speci-mens from the wells and water fountainwere cultured for fecal coliforms by theBureau of Laboratories, Department ofEnvironmental Resources, Harrisburg, PA.

On November 5, 1980, a dye test wasconducted by the Department of Envi-ronmental Resources of the sewage dis-posal system at the hardware store to seeif sewage from the store toilet could con-taminate the well. The procedure usedwas the standard State of Pennsylvaniatesting procedure. Four tablespoons ofuranine dye were flushed down the toiletin the hardware store. One week later,well water at the store was sampled andtested for presence of dye.


HARDWARE STORE WATER FOUNTAIN AND HEPATITIS A 697

In order to understand soil and rockconditions and possible surface and sub-surface water movements within thearea, a geologic report was obtained fromthe Environmental Geology Division,Bureau of Topographic and Geologic Sur-vey, Department of Natural Resources,State of Pennsylvania.

Case definition

A case was defined as a person witheither jaundice in the absence of gallbladder disease or previous blood transfu-sion plus elevation in bilirubin and one ofthe serum transaminases, or clinical ill-ness consisting of any four of the follow-

ing signs or symptoms: weakness, leth-argy or fatigue, anorexia, nausea or vom-iting, abdominal pain, or dark urine, oran ill person with less than four of theabove symptoms who had hepatitis Avirus IgM antibody (HAV IgM Ab).

RESULTS

Outbreak investigation

Endemic cases. From November 1979 toAugust 1980, 30 cases meeting our casedefinition had been reported from Lancas-ter County to the Pennsylvania StateHealth Department, for a biweekly rate of0 - 3 . As shown in figure 1, there was a

toUJCO<o

20-

19-

18-

17-

16-15-14-13-12-I I-10-

9-8-

7-

6-

5-

4-

3-2-

I -

ENDEMIC

EPIDEMIC

15 30 15 31NOV. DEC.

1979

fl15 29 15 31 15 30 15 31 15 30 15 31 15 31

FEB. MAR. APR. MAY JUN. JUL. AUG.

1980

ONSET (1/2 MONTH PERIODS)

15 30 15 31SEP. OCT.

FIGURE 1. Onsets of hepatitis A or infectious hepatitis cases in Lancaster County, Pennsylvania, fromNovember, 1979 to October, 1980 plotted by intervals of one-half month. Clear boxes or histograms are casesassociated with the epidemic at the hardware store. Stippled boxes are cases that occurred during the yearwhich were not associated with the hardware store and represent the background or endemic level oftransmission in the county.


698 BOWEN AND MCCARTHY

small peak in this background or "en-demic" level of transmission betweenJune 15 and July 15, 1980. This peak wasnot investigated for source. Greater than95 per cent of the cases reported beforeAugust 1980 were in Amish persons inthe area of the county north and east ofLancaster city. There was a slight pre-ponderance of male endemic cases (maleto female ratio 1.3:1) and 40 per cent ofthe endemic cases were < 16 years of age(table 1). The population of the county isabout 360,000; the crude annual attackrate for the county was 8.3/100,000. Thepopulation of the townships in which thecases occurred was 72,000 by the 197.0 USCensus. The attack rate for these town-ships was 41.7/100,000.

Epidemic cases. In September 1980, thelarge increase in reported hepatitis A casesprompted the current investigation, andwe identified 55 cases which met our casedefinition between August 1 and October31, 1980. Forty-nine cases in LancasterCounty were epidemiologically associatedwith the hardware store or the adjacentfamily home. Five cases in nearby south-ern Lebanon County attended a commonsocial gathering, but were not associated

with the hardware store. One case wasunrelated to either of these outbreaks.

The epidemic curve for this outbreak isshown in figure 2. The index case becameill on August 11 and the last case becameill on October 16. The outbreak had anearly peak between September 4 and Sep-tember 6 and a later peak from September13-September 24.

Over 90 per cent of the cases lived inLititz, Ephrata, or on farms in the ruralarea within a 5 mi (8.1 km) radius of ahardware store located 3 mi (4.8 km)southwest of Lititz, Clay Township, Lan-caster County. Six cases actually lived inEphrata while seven lived in Lititz. Bothtowns have municipal water supplies. Thefarms were served by wells.

The age and sex distribution of theepidemic cases is shown in the right handside of table 1. The male-to-female sexratio of the cases was 6:1. Forty of the 49cases (82 per cent) were between 16 and50 years of age. Only seven of theepidemic cases (14 per cent) were in per-sons <15 years of age. This was an un-usual age distribution for hepatitis A inLancaster County; as shown in table 1, 40per cent of the background or endemic

TABLE 1

Age and sex distribution of hepatitis A cases, Lancaster County, Pennsylvania, 1979—1980

Age(years)

0-56-10

11-15

16-2021-2526-30

31-4041-5051 +

Total

Male

006

en

to t

o

101

17

Endemic cases*Female

141

021

301

13

Total

147

246

402

30

Male

006

1284

732

42

Epidemic casestFemale

001

401

010

7

Total

007

1685

742

49

* Endemic cases include all hepatitis A cases reported from Lancaster County for one year between No-vember, 1979 and October, 1980 which were not associated with the epidemic at the hardware store. Thesecases represent the background level of transmission.

t Epidemic cases include only those cases associated with the hardware store from August—October, 1980.



12-,

I I -

10-

9-

8-

STORE CUSTOMER

STORE EMPLOYEE

CHOIR MEMBER

13 16 19 22 25 28

AUG.

3 6 9 12 15 18 21 24

OCT.

6 9 12 15 18 21 24 27 30

SEP.

ONSET (3-DAY PERIODS)

FIGURE 2. Epidemic curve by 3-day intervals of cases of hepatitis A in Lancaster County, Pennsylvania,from August to October, 1980. Clear boxes or histograms are store customers or family members. Lightstippled boxes are hardware store employees. Dark stippled boxes are youth choir members.

cases were less than 16 years of age. Theattack rate for the eight townships inwhich cases occurred was 106.6/100,000.

Occupations. The occupations of the 49cases are shown in table 2. Twelve (24.5per cent) were full-time or part-timeemployees of the hardware store, which

TABLE 2

Occupations of hepatitis A cases, Lancaster County,Pennsylvania, August —October, 1980

Occupation

Hardware store employeeFarmer/agriculturePart-time employee of

hardware storeConstruction (carpenter,

cabinetmaker, roofer, etc.)Mechanic (engine repair/power

equipment repair)Family of hardware store

ownerYouth choir memberCrop sprayerHousewife (friend of family of

owner of hardware store)Feed mill workerEgg processingTruck driverStudent

Total

No. ofcases

119

1

6

3

361

1111

_5

49

sold supplies and had an attached weldingshop where equipment including farmequipment was repaired. Nine cases werefarmers or agricultural workers whotraded at the hardware store; six wereemployed in construction; all did businesswith and had visited the hardware store.Three people were mechanics or repairedengines or power equipment and all threehad visited the hardware store. Threecases were family members of the ownerof the hardware store and lived in the ad-jacent home. One case was a family friendwho frequently visited the house. Fourpeople had other occupations (see table 2)which brought them in person to trade atthe hardware store. Five cases were stu-dents who had visited the home orhardware store.

Six people were members of a youthchoir, which met every other week at thehome of the owner of the hardware store.Four of these cases did not visit thehardware store on other business. Threecases were members of the choir and alsowere among a group of 11 females whoattended a slumber party at the home ofthe hardware store owner on August 23.The choir met at the family home on Au-gust 6 and 20 and September 3.



The index case was a 29-year-old maleemployee of the hardware store. The othertwo August cases and five of the sevencases in the first September peak werehardware store employees (figure 2).There was one late case in a storeemployee whose date of onset was October5. This person was out of town during Au-gust and did not return to work until Sep-tember 5. The second peak which occurredbetween September 14 and 25 was madeup principally of hardware store custom-ers, but also included members of theyouth choir.

Investigation into other possible sources.We carefully searched for other sourcesfor this outbreak; the 49 cases had onlyone thing in common—visiting or work-ing at the hardware store or visiting thefamily home next door in August or Sep-tember. The ill people did not share thesame municipal water supply. The major-ity lived on farms with wells, while 13cases were split equally between Lititzand Ephrata, which had separate watersystems. They did not attend a common so-cial group, church, or other organization.They had not travelled together. They didnot buy food at the same grocery store,bakery, or dairy. They did not use thesame medications or drugs. They did noteat at a common restaurant, fast food es-tablishment, or bakery. With the excep-tion of working at the hardware store, thecases did not work at the same place.Thus, the hardware store and adjacenthome were strongly implicated as thesource of the outbreak.

Investigation into the mode of transmis-sion. Once it had been established thatthe hardware store and adjoining homewere undoubtedly the source of the out-break, the mode of transmission was in-vestigated. No food or drinks were servedat the hardware store. There were novending machines of any kind. Theemployees brought lunch from home inbags or went home for lunch. They did notshare food, cigarettes, or drinks. The

employees did not have intimate personalor social contact with each other or withstore customers, making person-to-persontransmission unlikely. The hardwarestore and family home, but no otherhomes, were supplied by a common well.The hardware store contained a waterfountain, which was used by employeesand customers alike. A well, located inthe basement of the hardware store,supplied the water fountain and the waterfor the family home. There was a toiletfacility in the office of the store, whichwas used by all store employees includingthe index case, but not the store cus-tomers.

Two cases each occurred in only fourfamilies. They occurred 3-8 days apart.This made person-to-person transmissionunlikely. There were also clusters of casesin store employees between September 1and 6, in choir members between Sep-tember 15 and 24, and in customers whohad no intimate social contact with eachother or with store employees betweenSeptember 13 and 24 (see figure 2). Theseclusters were most likely common sourceclusters, not person-to-person transmittedchains of transmission.

We obtained water consumption his-tories from 44 store customers and 26youth choir members. Seven of eight storecustomers who drank water from thewater fountain became ill. Two of the 36customers who stated they did not drinkwater at the store were ill. Drinkingwater was highly associated with illness(p < 0.001 by Fisher's exact test). A simi-lar association between illness and drink-ing water was shown among the choirmembers. Twenty-six persons attendedboth the August 6 and August 20 choirmeetings; six out of 10 persons who drankthe water became ill versus none out of 16persons who did not drink the water (p <0.001). Thus, drinking water at the store orhome was highly associated with illness.

Further evidence that water was themode of transmission in this outbreak is



presented in table 3, which shows a"dose-response" relationship between de-gree of exposure to the water at the storeor family home and illness. The attackrates for the various groups of people atrisk during the outbreak decline as thelength and frequency of contact withwater at the store-home premises de-creases. Persons shown to be immune byserological testing (IgG but no IgM anti-body present in serum 4 - 8 weeks afteronset of illness) were excluded from thedenominators of at risk store employees,family members, youth choir members,and attendees of the pajama party. Theattack rate was 100 per cent for suscepti-ble family members who day and nightthroughout the outbreak drank waterfrom the well. The attack rate was 92 percent for susceptible store employees whohad daily exposure to the water, 27 percent for girls who attended the overnightpajama party at the family home in addi-tion to attending youth choir practice atthe home, and 9.6 per cent for other youthchoir members. A true attack rate could

not be obtained for store customers be-cause a complete listing of customers wasnot available, but a conservative estimateis that 1000 to 1500 customers were atrisk. Since there were 29 cases in storecustomers, an estimated attack rate forall customers would be 1.9 to 2.9 per cent.

Serology. Blood specimens were ob-tained from 32 clinical cases (two-thirdsof the cases) and 36 well persons in theyouth choir. All clinically ill persons hadIgM antibody to hepatitis A. Familymembers, store employees, and youthchoir members were bled, including thosenot ill. No asymptomatic recent infectionswere discovered among the youth choir,employees, or family members. Three ofthe 36 well youth choir members had IgGbut no IgM antibody, indicating that theyhad previously been infected. The anti-body prevalence for this age group was 8per cent.

Environmental investigation

The water supply for the hardwarestore and family home was a 200 ft (61 m)

TABLE 3

Attack rate of hepatitis A in five groups at risk, by exposure to contaminated well water,Lancaster County, Pennsylvania, August—October, 1980*

Group exposed

Household familymembers (bled)

Hardware storeemployees! (bled)

Attended slumber partyand youth choirmeetings (bled)

Youth choir memberswho did not attendslumber party (bled)

Hardware storecustomers (not bled)

No. of cases/no. susceptible

at risk

3/3

11/12

3/11

3/33

29/1000-1500§

No. immune

2

0

0

3

NT1

Attack ratet (%)

100

92

27

9

1.9-2.9

* Immune persons who had pre-existing IgG antibody were excluded.t Cases/at risk persons who were serologically susceptible.t One ill employee was not bled; full-time employees only.§ Denominator is estimate only.11 NT, not tested.



well drilled through the Beekmantowngroup of the carbonate rocks dolomite andlimestone. The water table level was 175ft (53 m). The rock is fractured and solu-ble in groundwater. Rain with any sur-face contaminants moves through the soiland easily through the rock along frac-ture lines down gravity gradients to lowerelevations.

Water is transferred from the well to an80 gall (303 liter) pressure tank by meansof a pump. Water passes from the pres-sure tank to the drinking fountainthrough a cartridge carbon filter. Twosamples of water were taken from thedrinking fountain on September 9, 1980.Both contained total coliforms of >200/100 ml and fecal coliforms of >60/100 ml.On September 22, two water samplesfrom the pressure tank of the well con-tained >200 fecal coliforms/100 ml.Water taken September 22 from wells lo-cated on two adjacent properties werenegative for fecal coliforms.

There was an employee toilet facilitylocated adjacent to the office of thehardware store. The septic tank and drainfield for the store toilet are located on flatground behind the store to the north (see

figure 3). The family home was located300 ft (91 m) to the east of the hardwarestore, and the septic tank and drain fieldfor the house toilet were behind the houseto the north. On November 5, four ta-blespoons of uranine dye were flusheddown the toilet in the hardware store. OnNovember 12, the well and water lineleading to the water fountain were sam-pled; no dye was measurable.

Control of the outbreak

On September 12, a bottled water sys-tem was substituted for well water at thestore and family home. Immune serumglobulin was administered to familymembers of cases. No cases were recog-nized after October 16. This time (fiveweeks) corresponds to one incubation pe-riod of hepatitis A.

Telephone interviews were conductedwith 39 cases four to six weeks after onsetof illness. These cases had 179 householdcontacts. No cases of jaundice were dis-covered. Likewise, no one with symptomscompatible with anicteric hepatitis suchas fatigue, low grade fever, anorexia,weight loss, tea colored urine, etc., wasfound. One person with gastrointestinal

SepticTank

(welh

Hardware Store

125 ft. SepticTank

• 3 0 0 ft.

125 ft.

FamilyHouse

II

FIGURE 3. Schematic diagram of location of well, septic tanks, and drain fields associated with hepatitis Aoutbreak, Lancaster County, Pennsylvania, August to October, 1980. (1 foot = 0.3 m.)



symptoms only was found. Thus, thesecondary attack rate and rate of anic-teric hepatitis appeared to be low.

DISCUSSION

This outbreak is unusual in that it is acommon source drinking water-associatedoutbreak with a prolonged intermittent orcontinuous period of well and drinkingwater contamination. Five groups ofpeople "at risk" because of exposure to thecontaminated water had attack rateswhich showed a "dose-response" relation-ship between attack rate and length anddegree of exposure. The outbreak was alsounusual because the attack rate for per-sons with highest exposure to the waterwas quite high (92-100 per cent) com-pared to other common source outbreaks(e.g., 3, 4). It was also unusual that noasymptomatic cases were found amongthe groups of persons from whom bloodwas collected (employees, family, youthchoir); the proportion of jaundiced caseswas very high; and the clinical secondaryattack rate was low. The unusual male tofemale sex ratio among the cases was dueto the fact that the persons doing businesswith or employed by the hardware storewere mostly males.

Most of the cases in this outbreak weredocumented serologically. Both IgG andIgM antibody were measured for two-thirds of the cases and additional personspotentially exposed to contaminatedwater but not clinically ill. These testswere extremely useful to show that fivepersons at risk were really already im-mune and that no asymptomatic infec-tions occurred. We believe that the use-fulness of serologic testing during hepa-titis A outbreak investigations has beenclearly established, and we strongly rec-ommend that these tests be carried outas part of future investigations.

It seems clear that the drinking waterrather than food, beverage, commongathering or person-to-person spread wasthe mode of transmission. No convincing

evidence for the other modes of transmis-sion could be obtained, although the pos-sibility that there could have been othermodes was carefully examined. All thecases had contact with the hardware storeor family home, but had nothing else incommon. Convincing evidence for wateras the source came from interviewing thepatients and controls among store cus-tomers and youth choir members; drink-ing water at the home or store was highlyassociated with illness. In addition, adose-response relationship between de-gree of exposure to the water and attackrate was demonstrated for the five groupsof people at risk at the store and home.The well and water fountain were shownto be contaminated by fecal coliforms.Lastly, the outbreak ceased one incuba-tion period after bottled water was substi-tuted for well water at the store andhome.

The route of contamination of the wellthat supplied the family home and thestore water fountain was not satisfactor-ily demonstrated. Because of the proxim-ity of the well to the store toilet and theseptic systems for the house and store,this route is highly suspect. The dye testperformed after the outbreak did not showthis route of contamination to be presentat the time the test was performed. Alter-natively, another surface water sourcecould have been responsible, althoughother nearby wells were not affected.

If the store toilet were the source of thewell contamination for most of the cases,then the period of possible well contami-nation was about 40 days, from about Au-gust 4 to September 12, when the bottledwater was substituted. This is so becausethe index case who used the store toiletbecame ill on August 11. If he excretedvirus from one week before onset of illness(6), he began shedding virus into thetoilet facility and, perhaps, the well onAugust 4. Other employees using thetoilet before their illnesses could alsohave contributed virus to contaminate the



well. Virus had to have been present aslate as September 5, when the last illstore employee returned to work after amonth's absence. Fecal contamination ofthe well was demonstrated on September9 and 22. One male choir member whoseillness began on September 16 attendedchoir practice at the home on August 6and 20 and September 3, but did not visitthe store or home at any other time. Hisexposure was probably on August 20. Thethree ill females who attended the pajamaparty on August 23 and the three choirpractices became ill September 17, 19,and 24. They were probably exposed onAugust 20 or 23. Other male choir mem-bers visited the store or home on other,unspecified, dates and may have been ex-posed at any one of the visits.

It is not clear whether virus was pres-ent continuously or intermittently in thewell water. The shape of the epidemiccurve with several peaks implies that,perhaps, more virus was present in thewater at some times than others or thatcontamination may have been intermit-tent.

The maximum attack rate in this out-break for employees and family memberswho had earliest and most prolonged ex-posure to the water was 93 per cent. Theattack rate for these maximally exposedpeople is high compared to other hepatitisA common source outbreaks reported inthe literature. In most outbreaks, attackrates have ranged from 7 — 8 per cent to 64per cent (3, 4, 7—9). Attack rates havebeen reported to be as high as 67—83 percent in residents of institutions and inchimpanzee associated outbreaks (10-12). The Holy Cross football team out-break (2) had an attack rate of >90 percent, but many cases were asymptomatic.

The absence of asymptomatic cases inthe relatively highly exposed youth choirgroup and in the attendees of the pajamaparty was unexpected, as was the rela-tively high proportion of jaundiced pa-tients (85 per cent). No evidence for clini-

cally compatible anicteric illness in ex-posed groups or families of cases was ob-tained. Osterholm et al. (8) did report a 4:1clinical to subclinical case ratio during afoodborne outbreak, but equal numbers ofasymptomatic and symptomatic caseswere reported by Rakela et al. (10). In theHoly Cross outbreak (2), 40 per cent ofinfections were asymptomatic in a compa-rable age group. Recent studies by Storchet al. (13) and Haider et al. (14) indicatethat young children are more likely tohave asymptomatic infections than adults.In these studies, symptomatic adults wereday-care center employees, parents, oradult household contacts of young chil-dren who attended day-care centers. Per-haps gross contamination of the well waterwith feces or water containing virus inthe Lancaster County outbreak resultedin clinical cases in those exposed and nolow dose exposures leading to asympto-matic infection.

Waterborne hepatitis A outbreaks havebeen reported many times in the past. In1966, Taylor et al. (15) analyzed the dataof 48 domestic and foreign outbreaks ofhepatitis; 17 of these outbreaks were as-sociated with wells, resulting in 1146cases (average, 67 cases/outbreak). TheLancaster County outbreak, with 49cases, is therefore slightly smaller thanthe average for well-associated wa-terborne outbreaks (15). The large wa-terborne outbreaks which have involvedhundreds or thousands of cases have beenassociated with contaminated municipalwater supplies.

During this investigation, more thanhalf of the cases were discovered as a re-sult of asking known patients if theyknew of other persons who were ill withthe same syndrome. This method of find-ing ill people should not be overlookedduring the epidemiologic investigation ofany outbreak, especially in rural areasand small towns or among particular so-cial, religious, or ethnic groups in whichthe members know each other well.



REFERENCES

1. Mosley JW, Schrack WD, Matter LD, et al. In-fectious hepatitis in Clearfield County,Pennsylvania. I. A probable waterborneepidemic. Am J Med 1959;27:555-68.

2. Morse LJ, Bryan JA, Hurley JP, et al. The HolyCross College football team hepatitis outbreak.JAMA 1972;219:706-8.

3. Eisenstein AB, Aach RD, Jacobsohn W, et al. Anepidemic of infectious hepatitis in a generalhospital. JAMA 1963;185:171-4.

4. Philip JR, Hamilton TP, Albert TJ, et al. Infec-tious hepatitis outbreak with Mai Tai as the ve-hicle of transmission. Am J Epidemiol1973;97:50-5.

5. Bradley DW, Fields HA, McCaustland KA, et al.Serodiagnosis of viral hepatitis A by a modifiedcompetitive binding radioimmunoassay for im-munoglobulin M anti-hepatitis A virus. J ClinMicrobiol 1979;9:120-7.

6. Bradley DW, Gravelle CR, Cook EH, et al. Cy-clic excretion of hepatitis A virus in experimen-tally infected chimpanzees. Biophysical charac-terization of the associated HAV particles. JMed Virol 1977;l:133-8.

7. Gavan DT, Nutt JW, An epidemic of waterborneinfectious hepatitis in France. Arch EnvironHealth 1970;20:523-32.

8. Osterholm MT, Kantor RJ, Bradley DW, et al.Immunoglobulin M-specific serologic testing inan outbreak of foodborne viral hepatitis, type A.Am J Epidemiol 1980;112:8-16.

9. Ruddy SJ, Johnson RF, Mosley JW, et al. Anepidemic of clam associated hepatitis. JAMA1969;208:649-55.

10. Rakela JR, Nugent E, Mosley JW. Viralhepatitis; enzyme assays and serologic proce-dures in the study of an epidemic. Am JEpidemiol 1977;106:493-501.

11. Hinthorn OR, Foster MT, Bruce HL, et al. Anoutbreak of chimpanzee associated hepatitis. JOccup Med 1974,16:388-91.

12. Mosley JW, Reinhardt HP, Hassler FT. Chim-panzee associated hepatitis. An outbreak inOklahoma. JAMA 1967;199:105-10.

13. Storch GS, McFarland LM, Velso K, et al. Viralhepatitis associated with day care centers.JAMA 1979;243:1514-18.

14. Hadler SC, Webster HM, Erben JJ , et al.Hepatitis A in a day care center. A communi-tywide assessment. N Engl J Med 1980;261:729-34.

15. Taylor FB, Eagen JH, Smith HFO, et al. Thecase for waterborne infectious hepatitis. Am JPublic Health 1966;56:2093-105.


Documents

HEPATITIS A ASSOCIATED WITH A HARDWARE STORE WATER … · 2017. 10. 3. · hardware store premises in August and September, 1980. These groups included hardware store employees, family