Evaluation of the clinical efficacy of prokinetic drugs inthe management of post-operative ileus: Can retrospective

data help us?

M.A. Smith a,*, G.B. Edwards a, B.L. Dallap b, P.J. Cripps a, C.J. Proudman a

a Equine Division, Department of Veterinary Clinical Science and Animal Husbandry, University of Liverpool, Leahurst,

Neston, Wirral, Liverpool CH64 7TE, UKb University of Pennsylvania, New Bolton Center, 382 West Street Road, Kennett Square, PA 19348-1692, USA

Accepted 27 June 2004

Abstract

The objectives of the study were to determine whether retrospective data can be used to answer questions about the efficacy of

prokinetic agents when used to treat horses with post-operative ileus (POI). We describe prevalence and mortality of POI with ref-

erence to treatment with four prokinetic agents. By combining data from two Hospitals a study population of 55 horses with POI

following pedunculated lipoma obstruction (PLO) was established. Univariable and multivariable associations were determined

between short term survival and potential explanatory variables.

With death as outcome in multivariable models, breed and hospital were significantly associated with outcome but the use of

prokinetic agents was not (P = 0.15). However, sample size estimates indicate the low power of this study to detect differences in

outcome. It was not possible definitively to evaluate the efficacy of prokinetics as treatment for POI following PLO using retrospec-

tive data. The data were suggestive of limited efficacy of prokinetics as treatment for POI. It is postulated that the identified asso-

ciation between hospital and survival reflects differences in clinician decision making. The study highlights the need for further

prospective studies using randomised clinical trials to evaluate accurately the efficacy of prokinetic agents. This report illustrates

difficulties with performing retrospective analysis of clinical data to determine the efficacy of treatment regimes.

� 2004 Elsevier Ltd. All rights reserved.

Keywords: Pedunculated lipoma; Prokinetics; Ileus; Evidence based medicine

1. Introduction

The aetiopathogenesis of post-operative ileus (POI) is

only partly understood. The condition has been linked

with endotoxaemia, bowel distension, peritonitis, hypo-kalemia, anaesthetic agents and analgesics (Gerring and

Hunt, 1986; King and Gerring, 1988, 1991; Lester et al.,

1998; Merritt et al., 1998). Edwards and Hunt (1986)

first documented a 16.2% prevalence of ileus in all cases

undergoing gastrointestinal surgery over a 15 year

period. More recently, prevalences of 3% (Philips

and Walmsley, 1991), 21% (Blikslager et al., 1994) and

9.6% (Proudman et al., 2002) have been reported, and

a prevalence of 9% in a population of 63 horses under-going surgery for small intestinal lesions only has been

cited (Freeman et al., 2000).

It has recently been shown that horses suffering

from pedunculated lipoma obstruction (PLO) are at in-

creased risk of developing POI, being three times more

likely to suffer from POI compared with horses suffer-

ing from other intestinal pathology, odds ratio 3.2

(95% CI 1.3, 7.5) (French et al., 2002). We postulate

1090-0233/$ - see front matter � 2004 Elsevier Ltd. All rights reserved.

doi:10.1016/j.tvjl.2004.06.006

* Corresponding author. Tel.: +44 151 794 6041; fax: +44 151 794

6034.

E-mail address: m.a.smith@liv.ac.uk (M.A. Smith).

www.elsevier.com/locate/tvjl

The Veterinary Journal 170 (2005) 230–236

TheVeterinary Journal

that the PLO population is therefore a good clinical

model for evaluating the efficacy of prokinetics in treat-

ing POI.

Pharmacological studies have demonstrated that

dopaminergic and adrenergic hyperactivity may have a

key role in initiating POI. A number of studies providea pharmacological rationale for using metaclopramide

as an adrenergic and dopamine antagonist and cisapride

as an indirect cholinergic agent for the treatment of POI

(Hunt and Gerring, 1986; Gerring and Hunt, 1986;

Gerring and king, 1989). More recent research suggests

that the prokinetic action of metaclopramide is more

likely due to its action as a 5HT4 agonist (Doherty

and Frazier, 1998).Erythromycin has been shown to increase solid phase

gastric emptying significantly and is thought to exert

prokinetic effects via activation of motilin receptors

(Ringger et al., 1996) although other studies have shown

little effect of erythromycin on gastric emptying (Doh-

erty and Frazier, 1998) or intestinal motility (Roussel

et al., 2000) and the drug is reported to be ineffective

in man (Bonacini et al., 1993). The amide anaestheticlignocaine (lidocaine) is also administered as a proki-

netic agent, however the mechanism of action of this

drug has yet to be elucidated (Groudine et al., 1998).

None of the intestinal prokinetic agents mentioned

above has been subject to rigorous clinical efficacy trials.

Their use in horses suffering POI is based on personal

experience of clinicians and anecdote.

The aim of our study was to establish whether retro-spective clinical data can be used to evaluate the efficacy

of prokinetic drugs used in the treatment of POI. A sin-

gle category of surgical colic was chosen in order to

standardise the initiating insult and the population of

horses studied.

2. Materials and methods

The databases of two large equine referral centres

were reviewed and 81 horses diagnosed with PLO affect-

ing the small intestine identified from medical records at

Hospital 1 and 179 horses identified at Hospital 2 over

�15 and 10 year periods, respectively. Diagnosis in all

cases was confirmed during surgery or at post-mortem.

Two hundred and two horses taken to surgery wereidentified, 55 of which went on to develop POI.

Records were reviewed and a number of variables ex-

tracted for each horse. Signalment, heart rate, PCV,

amount of reflux obtained at admission and duration

of colic prior to presentation were recorded; the length

of intestine resected, anastomosis type performed,

outcome of surgery, time to first feed, duration of hospi-

talisation, occurrence of ileus and duration of post-operative reflux were ascertained where applicable.

The use of one or more of four prokinetic agents; meta-

clopramide, cisapride, erythromycin and lignocaine/

lidocaine was determined. The variable ‘‘prokinetics

Y/N’’ (Tables 3 and 4) included the following pharma-

cological agents: metaclopramide (n = 12), cisapride

(n = 5), lignocaine/lidocaine (n = 1), erythromycin (n =

2) and a combination of the four (n = 10). Twenty-fiveof 55 horses with POI did not receive a prokinetic agent.

Short term survival was defined as survival to discharge

from the hospital.

Post-operative ileus was defined as obtaining two or

more litres of gastric reflux on more than one occasion

when a nasogastric tube was passed. Post-operative sur-

vivors that were diagnosed as suffering from ileus were

identified. In addition, horses suffering ileus that sur-vived beyond three days post-operatively were identi-

fied. Horses that were later found to have ileus as a

result of a mechanical obstruction confirmed at a second

surgery or at post-mortem were excluded from the POI

population.

The following treatment regimes were used; at Hospi-

tal 1, cisapride was administered per rectum, in solution

with 5 mL DMSO at a dose of between 100 and 200 mg,q.i.d. according to the clinicians� discretion. Metaclopra-

mide was administered as 30 mg in 1 l of polyionic fluids

(0.06 mg/kg) administered as an intravenous (IV) infu-

sion over 1 h, q.i.d.

Erythromycin was administered as 1 g in 1 L of sterile

saline, given as an infusion over 1 h, q.i.d. At Hospital 2,

cisapride was administered intramuscularly (IM), at a

dose rate of 0.1 mg/kg TID and metaclopramide wasadministered by IV infusion at a rate of 0.04 mg/kg

every hour. Erythromycin was not used at Hospital 2.

Both hospitals administered lignocaine/lidocaine IV as

a loading bolus of 1.3 mg/kg given in polyionic fluids

over 15 min, followed by a maintenance infusion of

0.05 mg/kg/min. The choice of prokinetic used varied

according to the clinicians� preference. Standard post-

operative treatment protocols for the use of IV fluidtherapy, antibiotics and nonsteroidal anti-inflammatory

drugs were used in each hospital (White, 1990).

2.1. Statistical analysis

A sample size estimate was performed based on a lo-

gistic regression model and using the Egret Siz pro-

gramme in order to determine the power of the study,and to establish what size odds ratio our cohort would

allow us to detect. A single variable (use or not of proki-

netics) was used, assuming that 66% of cases survived,

with 50% of all horses in POI (n = 55) receiving proki-

netics. A critical probability of P < 0.05 was used. Our

study population of 55 horses with POI gave 80% power

for the detection of an odds ratio of >5 under the con-

ditions stated.Univariable analyses of data were performed using

Kruskal–Wallis or v2 tests to detect differences between

M.A. Smith et al. / The Veterinary Journal 170 (2005) 230–236 231

the two hospital populations, and between groups of

horses within each population. Continuous variables

were categorised using biologically justifiable cut-points

as determined from preliminary analyses. For example,

a step function was noted for heart rate at 66 bpm and

so this cut off was used to convert this variable intocategories. Univariable associations with survival for

55 horses suffering from POI were determined using lo-

gistic regression (EGRET). Variables demonstrating

some evidence of association with outcome (P < 0.2)

were carried forward into multivariable models. Multi-

ple logistic regression analysis with backwards elimina-

tion was performed to evaluate multivariable

relationships with death as the outcome of interestand to control for potential confounding variables.

Variables with a likelihood ratio test statistic P value

of <0.05 were considered to be significantly associated

with outcome. We explored the use of random effects

terms to control for variability due to surgeon or to

year of surgery. Neither term had any influence on

the final multivariable model, likely due to the small

sample size. Biologically plausible interaction termswere also analysed.

3. Results

3.1. Univariate analysis

Tables 1 and 2 indicate similarities and differences inthe PLO population at the two hospitals. Gender distri-

bution and median heart rate were similar between the

two hospitals. However, the median age of PLO cases,

breed distribution, PCV at presentation, volume of re-

flux at presentation, duration of colic prior to presenta-

tion, numbers euthanased intra-operatively, numbers

recovered from surgery, prevalence of POI and survival

to discharge were significantly different. These findingssuggest that the two hospital populations were not

similar.

3.2. Risk factors for survival following POI

Table 3 shows univariable associations with survival

in POI cases from both hospitals. Sex, breed, resection

length, use of prokinetics and hospital all showed someevidence of association with survival to discharge

(P < 0.2). These five variables were carried forward into

a multivariable logistic regression model. Sex and resec-

tion length were discarded from the model due to poor

fit. Our final multivariable model (Table 4) indicated

that breed and hospital were significantly associated

with outcome, with Thoroughbreds being more likely

to survive than ponies (P = 0.01), and horses at Hospital2 being at greater risk of death than horses at Hospital 1

(P = 0.01). Use of prokinetics was not significantly asso-

ciated with outcome (P = 0.15, OR 0.32, 95%; CI 0.07–

1.50).

To explore further the association of hospital with

outcome, given that there were differences between hos-

pitals in the numbers of horses euthanased very early on

in the post-operative period (Table 2), a second model

was run using only those horses in POI which survived

beyond three days in order to rule out treatment bias.The rationale for this was: at Hospital 1 a greater pro-

portion of horses were subjected to intra-operative

euthanasia than at Hospital 2 and we hypothesised that

this removed the sickest horses from the post-operative

population at Hospital 1. Hospital 1 also had the highest

prevalence of prokinetic use. At Hospital 2 more horses

were euthanased in the early post-operative period, com-

monly without prokinetics being used. The three day cutoff point removed from the analysis sick horses that

were euthanased early in the post-operative period irre-

spective of the use of prokinetics. In this second model,

use of prokinetics was not significantly associated with

outcome (P = 0.43, OR = 0.50, 95% CI 0.09–2.78), how-

ever Hospital was (P = 0.02, OR = 16.4, 95% CI 1.65–

163.86).

An interaction term between breed and the use ofprokinetics was not significant (P = 0.98). We were

Table 1

Inter-Hospital comparison of clinical data at presentation of 260

horses with PLO

Variable Hospital 1

(n = 81)

Hospital 2

(n = 179)

P-value

Gender

M 24 64

G 55 109

S 2 6 0.67

Breed

TB/TBX 29 36

Cob 3 26

Arab/AX 10 17

Pony 5 68

QH 17 0

Other 17 10

Unknown 0 22 <0.01

Median age (years) 19 17 <0.01

Median heart rate

at presentation

60 bpm 60 bpm 0.81

Median PCV at

presentation

39% 43% 0.05

Median volume of reflux

at presentation

2 l 0 l <0.01

Duration of colic

prior to presentation

8 h 13 h <0.01

The null hypothesis was tested for each variable using v2 analysis forcategorical data and Kruskall–Wallis test for continuous data. A

critical probability of P < 0.05 was assumed. Key: M = mares,

G = geldings, S = stallions, AX = Arab cross, TB = thoroughbred,

TBX = thoroughbred cross, QH = quarter horse, PCV = packed cell

volume.

232 M.A. Smith et al. / The Veterinary Journal 170 (2005) 230–236

Table 2

Outcome of surgery for 249 horses with PLO that underwent general anaesthesia

Hospital 1 (n = 70) Hospital 2 (n = 179) P-value

Euthanased intra-operatively 18 (22%) 19 (10.5%) 0.01

Euthanased in recovery 2 (3%) 8 (4.5%) 0.44

Recovered from surgery 50 (62%) 152 (85%) <0.01

Prevalence of ileus 20 (40%) 35 (23%) 0.01

Mean time to first feed (h) 48 53 0.85

Duration of hospitalisation (days) 10.5 (median) 9 (median) 0.11

Duration of post-op reflux (h) 30 (median) 36 (median) 0.47

Survival to discharge 48 (59%) 119 (66.5%) 0.03

The null hypothesis was tested for each variable using v2 analysis for categorical data and Kruskall–Wallis test for continuous data. A critical

probability of P < 0.05 was assumed.

Table 3

Univariable associations with survival for 55 horses suffering from POI (continuous variables categorised)

Variable Survived n Died n Odds ratio 95% Confidence interval v2 P-value

Age

618 20 13 1

>18 14 8 0.88 0.30–2.68 0.82

Sex

Female 10 10 1

Male 24 11 0.46 0.15–1.42 0.18

Breed

Pony 3 9 1

TB/TBx 13 2 0.05 0.01–0.37

QH 6 1 0.06 0.01–0.67

Other 12 7 0.25 0.05–1.20 <0.01

HR at presentation

666 bpm 15 11 1

>66 bpm 15 10 0.91 0.30–2.78 0.87

PCV at presentation

645% 16 9 1

>45% 13 12 1.64 0.53–5.09 0.39

Duration of colic

611 h 15 5 1

>11 h 17 11 1.94 0.55–6.88 0.31

Reflux on admission

62 l 20 16 1

>2 l 13 5 0.48 0.14–1.63 0.24

Resection length

67 feet 20 8 1

>7 feet 13 12 2.31 0.74–7.18 0.15

Anastomosis type

Jej-jejunostomy 15 7 1

Jej-caecostomy 15 14 2 0.63–6.35 0.24

Prokinetics

No 11 14 1

Yes 23 7 0.24 0.08–0.76 0.01

Hospital

1 19 1 1

2 15 20 25.33 3.04–210.94 <0.01

Key: TB = thoroughbred; QH = quarter horse; HR = heart rate; PCV = packed cell volume.

M.A. Smith et al. / The Veterinary Journal 170 (2005) 230–236 233

unable to fit interaction terms between hospital and

prokinetics as no fatalities were recorded in hospital

one�s horses receiving prokinetics, thus resulting in zero

cells in this group.

4. Discussion

The population of horses with PLO was chosen for

this retrospective study because it allowed POI to be

studied in a group of horses with similar intestinal

pathology. In each case a definitive diagnosis could be

made and ischaemia of small intestine had probably oc-

curred, resulting in exposure to endotoxin, although to

varying degrees. Resection of small intestine was per-

formed in 139 (83%) of 167 horses that survived to dis-charge, supporting the assumption that irreversible

changes in intestinal viability were identified at surgery

in the majority of horses. A high correlation between

the incidence of POI and the presence of ischaemic gut

has been reported (King and Gerring, 1991). Approxi-

mately two thirds of both PLO populations were geld-

ings, and the condition affected older animals. These

findings are in agreement with a previous report (Ed-wards and Proudman, 1994). The consistent age and

sex distribution of this population of surgical colic pa-

tients over the past decade gives support to our selection

of this group as a stable group for evaluation of post-

operative treatment protocols.

The prevalence of POI following PLO at both hospi-

tals in this study (42% and 23% at Hospitals 1 and 2,

respectively) is considerably higher than values reportedin the literature for surgical correction of small intestinal

pathology in general (Blikslager et al., 1994 (21%); Ed-

wards and Hunt, 1986 (16%); Philips and Walmsley,

1991 (3%)). This is supported by findings in a recent

study of post-operative complications of colic surgery

where horses recovering from PLO were found to be

three times more likely to suffer POI than horses with

other lesions (French et al., 2002). This finding again

supports our use of this group of colic patients for eval-

uation of POI.

Data were collected from two large referral hospitals

in order to maximise study power, however sample size

estimates show that in spite of combining two large data

sets acquired over a long time period this study was onlyable to detect odds ratios of >5 with 95% confidence and

80% power. This shows clearly that the statistical power

of our study was low. Univariable analysis showed that

there were significant differences in age, breed, volume

of reflux at presentation and duration of colic prior to

presentation at the two hospitals indicating marked dif-

ferences between the two populations studied.

We have addressed the issue of confounding by thesevariables by the use of multivariable analysis. Any ben-

efit obtained by adding additional hospital populations

to the study in an attempt to increase power would

clearly be offset by increased variability between the hos-

pital populations, however this is the only means of

increasing study power using retrospective data. Inclu-

sion of more hospital populations may provide a better

representation of variation between them. We have goneon to combine the two populations in order to achieve

maximum study power with the data available, but have

had to accept limitations in interpretation of the results.

Our final model offered no evidence to support the asso-

ciation between prokinetic use and survival but low

study power limits the interpretation of this negative

finding.

To date, the only critical evaluation of pharmacolog-ical intervention in clinical cases with POI has been that

of Dart et al. (1996). This retrospective study demon-

strated reduced incidence and severity of POI in 70

horses undergoing small intestinal resection and anasto-

mosis when metaclopramide was given as a continuous

infusion. Metaclopramide was used as part of the rou-

tine post-operative management in 44 horses but differ-

ences in survival between treatment groups were notstatistically significant (Dart et al., 1996).

One case series reports that the use of cisapride post-

operatively in 42 colic patients was beneficial given that

POI was seen in only two horses (DeGeest et al., 1991),

however only 18/42 horses had identifiable small intesti-

nal pathology at surgery and only 10 appear to have suf-

fered from a strangulating lesion of small intestine. A

multicentre trial reported that of 81 horses treated withcisapride, POI was prevented in 57 cases, and the

authors concluded that the drug was highly efficacious

in preventing idiopathic ileus (Gerring et al., 1991).

Whilst all horses in the study had suffered small intesti-

nal pathology, only 43 horses suffered a strangulating le-

sion of small intestine. Cisapride was judged to be

ineffective in preventing POI in 7% of the horses, and

this value is similar to values reported for prevalenceof POI in populations undergoing small intestinal

surgery that do not receive prokinetics, for example

Table 4

Parameter estimates for final multivariable model for 55 horses

suffering from POI (outcome = death)

Variable Odds ratio 95% Confidence interval LRS P-value

Breed

Pony 1

TB/TBx 0.06 0.01–0.53 0.01

Other 0.22 0.03–1.41 0.11

Prokinetics

No 1

Yes 0.32 0.07–1.50 0.15

Hospital

1 1

2 28.12 2.65–298.50 0.01

LRS = likelihood ratio statistic.

234 M.A. Smith et al. / The Veterinary Journal 170 (2005) 230–236

Freeman et al. (2000) reported a prevalence of 9% in 63

horses undergoing small intestinal surgery.

Elective euthanasia was performed on more horses

intra-operatively at Hospital 1 compared with Hospi-

tal 2, however at Hospital 2 elective euthanasia was

performed more often in the early post-operative per-iod. This created a potential source of treatment bias

when combining the data. In order to remove the

leverage exerted by the sickest horses (those euthan-

ased in the early post-operative period) irrespective

of the use of prokinetics, a three day cut off point

was used. When trying to assess the effectiveness of

use of prokinetics in the post-operative period only

those horses surviving beyond three days post surgerywere analysed in the second multivariable model. In

this model the use of prokinetics was not significantly

associated with decreased risk of death (P = 0.43),

however Hospital remained significantly associated

with outcome, with horses at Hospital 2 remaining

at increased risk of death compared with those at

Hospital 1 (P = 0.02, OR = 16.4, 95%; CI 1.65–

163.86).We conclude that despite standardising the study

population by selection of one group of surgical colic

patients, maximising case numbers by combining data

from two large referral hospitals, and performing multi-

variable analysis to control for confounding, we are un-

able to answer the question regarding the efficacy of

prokinetics as treatment for POI following PLO using

our retrospective data. More hospitals and a corre-spondingly larger data set are needed before we can con-

clude that retrospective analysis is not useful to answer

our hypothesis. We believe that these data are suggestive

of limited efficacy of prokinetics as treatment for POI,

however there were too many different prokinetics used

and too many different treatment strategies used to be

able to draw a definite conclusion. We have identified

breed and hospital as major sources of variability inpost-operative survival and postulate that the associa-

tion between hospital and survival reflects differences

in clinician decision making between the two hospitals.

This report highlights some of the difficulties encoun-

tered when performing retrospective analysis of clinical

data, and emphasises the need for prospective clinical

trials in order to better assess efficacy of treatment with

prokinetic agents.

Acknowledgement

Meredith Smith was funded by the Home of Rest for

Horses. We thank colleagues at the George D. Widner

large Animal Hospital, University of Pennsylvania,

USA, and the Philip Leverhulme Large Animal Hospi-tal, University of Liverpool, U.K. The colic database

at the University of Liverpool has been supported by

the Home of Rest for Horses and PetPlan Charitable

Trust.

References

Blikslager, A.T., Bowman, K.F., Levine, J.F., Bristol, D.G., Roberts,

M.C., 1994. Evaluation of factors associated with post-operative

ileus in horses: 31 cases (1990–1992). Journal of the American

Veterinary Medical Association 205, 1748–1752.

Bonacini, M., Quiason, S., Reynolds, M., Gaddis, M., Pemberton, B.,

Smith, O., 1993. Effect of intravenous erythromycin on post-

operative ileus. American Journal of Gastroenterology 88, 208–

211.

Dart, A.J., Peauroi, J.R., Hodgson, D.R., Pascoe, J.R., 1996. Efficacy

of metaclopramide for treatment of ileus in horses following small

intestinal surgery: 70 cases (1989–1992). Australian Veterinary

Journal 74, 280–284.

DeGeest, J., Vlaminck, K., Muylle, E., Deprez, P., Sustronck, B.,

Picavet, M.T., 1991. A clinical study of cisapride in horses after

colic surgery. Equine Veterinary Education 3, 138–142.

Doherty, T.J., Frazier, D.L., 1998. Effect of intravenous lidocaine on

halothane minimum alveolar concentration in ponies. Equine

Veterinary Journal 30, 300–303.

Edwards, G.B., Hunt, J.M., 1986. An analysis of the incidence of

equine post-operative ileus (POI) and an assessment of the

predisposing factors. In: Proceedings of the Second Equine Colic

Research Symposium, pp. 307–312.

Edwards, G.B., Proudman, C.J., 1994. An analysis of 75 cases of

intestinal obstruction caused by pedunculated lipoma. Equine

Veterinary Journal 26, 18–21.

Freeman, D.E., Hammock, P., Baker, G.J., Goetz, T., Foreman, J.H.,

Schaeffer, D.J., Richter, R.A., 2000. Short and long-term survival

and prevalence of post-operative ileus after small intestinal surgery

in the horse. Equine Veterinary Journal Supplement 32, 42–51.

French, N.P., Smith, J.E., Edwards, G.B., Proudman, C.J., 2002.

Equine surgical colic: risk factors for post-operative complications.

Equine Veterinary Journal 34, 444–449.

Gerring, E.L., Hunt, J.M., 1986. Pathophysiology of equine post-

operative ileus: effect of adrenergic blockade, parasympathetic

stimulation and metaclopramide in an experimental model. Equine

Veterinary Journal 18, 249–255.

Gerring, E.L., King, J.N., 1989. Cisapride in the prophylaxis of equine

post-operative ileus. Equine Veterinary Journal Supplement 7, 52–

57.

Gerring, E.L., King, J.N., Edwards, G.B., Pearson, H., Walmsley, J.P.,

Greet, T.R.C., 1991. A multicentre trial of cisapride in the

prophylaxis of equine post-operative ileus. Equine Veterinary

Education 3, 143–145.

Groudine, S.B., Fisher, H.A.G., Kaufman, R.P., Patel, M.K., Wilkins,

L.J., Mehta, S.A., Lumb, P.D., 1998. Intravenous lidocaine speeds

the return of bowel function, decreases post-operative pain and

shortens Hospital stay in patients undergoing radical retropubic

prostatectomy. Anesthesia Analgesia 86, 235–239.

Hunt, J.M., Gerring, E.L., 1986. A preliminary study of the effects of

metaclopramide on equine gut activity. Journal of Veterinary

Pharmacology and Therapeutics 9, 109–112.

King, J.N., Gerring, E.L., 1988. Actions of the novel gastrointestinal

prokinetic agent cisapride on equine bowel motility. Journal of

Veterinary Pharmacology and Therapeutics 11, 314–321.

King, J.N., Gerring, E.L., 1991. The action of low dose

endotoxin on equine bowel motility. Equine Veterinary

Journal 23, 11–17.

Lester, G.D., Merritt, A.M., Neuwirth, L., Vetro-Widenhouse, T.,

Steible, C., Rice, B., 1998. Effect of erythromycin lactobionate

on myoelectric activity of ileum, cecum, and right ventral

M.A. Smith et al. / The Veterinary Journal 170 (2005) 230–236 235

colon, and cecal emptying of radiolabeled markers in clinically

normal ponies. American Journal of Veterinary Research 59,

328–334.

Merritt, A.M., Burrow, J.A., Hartless, C.S., 1998. Effect of xylazine,

detomidine and a combination of xylazine and butorphanol on

equine duodenal motility. American Journal of Veterinary

Research 59, 619–623.

Philips, T.J., Walmsley, J.P., 1991. Retrospective analysis of the results

of 151 exploratory laparotomies in horses with gastrointestinal

disease. Equine Veterinary Journal 25, 427–431.

Proudman, C.J., Smith, J.E., Edwards, G.B., French, N.P., 2002.

Long-term survival of equine surgical colic cases. Part 1: patterns

of mortality and morbidity. Equine Veterinary Journal 34, 432–

437.

Ringger, N.C., Lester, G.D., Neuwirth, L., Merritt, A.M., Vetro, T.,

Harrison, J., 1996. Effect of bethanecol or erythromycin on gastric

emptying in horses. American Journal of Veterinary Research 37,

1771–1775.

Roussel, A.J., Hooper, R.N., Cohen, N.D., Bye, A.D., Hicks, R.J.,

Bohl, T.W., 2000. Prokinetic effects of erythromycin on the ileum,

caecum, and pelvic flexure of horses during the post-operative

period. American Journal of Veterinary Research 61, 420–424.

White, N.A., 1990. In: White, N.A. (Ed.), The Equine Acute

Abdomen. Lea and Febiger, Philadelphia.

Book review

P. Durr, and A. Gatrell (Eds.), GIS and Spatial Analysis

in Veterinary Science, Wallingford, CABI Publishing,

2004, ISBN 0851996345, p. 320, £60

This multi-authored book came about as a result of

the first GISVet conference held at Lancaster University

in September 2001. Following the conference a special

issue of Preventive Veterinary Medicine (2002, Volume56, Issue 1) was published that included a selection of

the original research papers presented at the conference.

This book contains revised and expanded versions of

several of the papers delivered at the conference as well

as one additional invited contribution.

The book is divided into three parts that explore the

use of spatial statistical analysis and geographical infor-

mation systems (GIS) in a veterinary context. The firstsection contains two chapters that introduce the topic

and cover basic concepts and principles with veterinary

examples. Chapter 1, ‘‘The Tools of Spatial Epidemiol-

ogy: GIS, Spatial Analysis and Remote Sensing’’ gives

an excellent introduction to the topic and basic principles

with several easy to understand veterinary examples.

This chapter ends with an overview of the remaining

chapters in the book. Chapter 2, ‘‘Spatial Epidemiology

and Animal Disease: Introduction and Overview’’ extends

the concepts presented in the first chapter and applies

them to veterinary epidemiology using bovine spongi-

form encephalopathy, bovine tuberculosis and multi-

drug-resistant Salmonella newport as examples.

The second section also contains two chapters that

look at the broader context of GIS and spatial statistics

in biomedical and human public health fields. Chapter 3,

‘‘Geographical Information Science and Spatial Analysis

in Human Health: Parallels and Issues for Animal Health

Research’’ does exactly as the title suggests! Chapter 4,

‘‘Spatial Statistics in the Biomedical sciences: Future

Directions’’ discusses the application of advanced spatial

statistical methodology to health issues.

The third section contains six chapters that deal with

the application of GIS and spatial analysis to a varietyof areas in the veterinary field, including a general ap-

proach to animal health and disease surveillance, parasi-

tology, modelling the spatial and temporal spread of

animal diseases, companion animal epidemiology, epi-

demic disease response and the management of wildlife

diseases. These final chapters all contain excellent exam-

ples that illustrate the application presented in each

chapter.Each chapter includes a list of references and there is

also an appendix at the end of the book that provides an

excellent overview of GIS and spatial analysis resources.

The appendix contains information and advice on soft-

ware, obtaining spatial data and the GISVet Website

(http://www.gisvet.org). This book provides both a basic

introduction to the concepts as well as in-depth coverage

of a large breadth of topics that make it a valuable ref-erence for those with an interest in applying these meth-

ods in veterinary science.

Vicki J. Adams

Small Animal Epidemiologist, Animal Health Trust

Lanwades Park, Kentford, Newmarket

Suffolk CB8 7UU, UK

E-mail address: vicki.adams@aht.org.uk

doi:10.1016/j.tvjl.2005.02.021

236 M.A. Smith et al. / The Veterinary Journal 170 (2005) 230–236