Br. J. Surg. 1992. Vol. 79, October 1073-1 076
Comparative audit: an experimental study of 147 882
general surgical admissions during
I990
D.C. Dunn and S. Fowler Confidential Comparative Audit Service, The Royal College of Surgeons of England, 35-43 Lincoln’s Inn Fields, London WC2A 3PN, UK Correspondence to: Mr D. C. Dunn
In 1991,1025 general surgical Fellows of The Royal College of Surgeons of England were circulated with a pro forma and asked to submit local audit results f o r admissions during 1990 to a confidential comparative audit service. The individual topics of cholecystectomy and colorectal resection were studied. Data returned by 160 surgeons concerned 147882 admissions including 122 620 operations. Overall mortality rates ranged f r o m 0 to 5 per cent and morbidity rates f r o m 0 to 22 per cent. Laparoscopic cholecystectomy was associated with one-quarter of the mortality rate and two-thirds the morbidity rate of open cholecystectomy. Of the 33 surgeons who responded to a survey after the presentation of results, all wished to continue the exercise in future years; 39 per cent had been stimulated to perform further analyses and 15 per cent had changed practice habits as a result. Comparative audit involving large numbers of patients and surgeons is feasible and seems benejicial to participants.
Surgeons in the UK are required to audit practice. The ability to d o this varies and is often dependent on the availability of suitable information technology’.’. When information is available an understanding of local practice may be reached3. Results may then need to be compared within a larger group. This stimulus produced comparative audit and provided the opportunity to define what is required of information systems4. Comparative audit deals with ‘raw’ data, the collection of which is largely uncontrolled. Unevenness in data collection is, however, compensated for by the large numbers of patients involved. Information from such databases can then be used for comparison with unit or personal results. If investigation indicates that poor practice is the cause of difference, corrective action can be instituted. Comparison of individual working conditions can be made, giving a greater insight into the special characteristics of local practice. A confidential comparative audit service was set up at The Royal College of Surgeons of England and a pilot meeting organized for general surgeons to develop, test and evaluate the methodology.
Materials and methods
and surgical workload analysis were those a g r x d by a working party’ and are shown in Table 2. Surgeons graded their operations, where possible, into minor, intermediate and major according to the British United Provident Association classification6 and this grading was converted into the intermediate procedure equivalent value’ (Tuble 3). This enabled a weighted value to be produced for operative workload.
Table 1 Data requatcd
it7
tilt pro formu srtii to cotisultat~t.s
Resources Manpower. beds, theatre sessions Workload Admissions, emergencies, procedures, stay nights Case mix Age groups, diagnoses, operative groups Complications Mortality, complication grades and types Topics Cholecystectomy Numbers, deaths, complications Colorectal resection Numbers, deaths, complications
A booklet pro forma was designed and circulated to all general surgical Fellows of The Royal College of Surgeons of England working in England and Wales. Information was requested on activity in 1990; a summary of the items requested is shown in Table I . The pro forma wascirculated to 1025 surgeons at the end of March 1991 and completed returns requested by the end of April 1991. This time scale meant that only those with access to the data requested were likely to take part. Assurance was given that the data would be treated confidentially. A confidential number was allocated and all further handling of the data performed in connection with this number only. Some members of the secretariat had access to the names of the surgeons but no surgeon was given this information. Participants were asked to enter ‘n.a.’ where data were not available and ‘0’ for a nil return. Returned data were entered into a spreadsheet (Super Calc 5; Computer Associates, San Jose, California, USA). Each data item was ranked and means and 25.50 and 75 percentiles produced. Sheets were produced for each confidential number and the rankings and positions on each of the forty analyses could be seen. An example chart is shown in Figure I . Analyses were made of the manpower available to consultants. Recommended manpower equivalent values for use in surgical audit
0007-1323/92/101073~ 0 1992 Butterworth-Heinemann Ltd
1073
Audit of general surgical admissions: D. C. Dunn and S. Fowler Table 2 Recommended d u e s for use in surgical audit and surgical workload a n a l l s i . ~ ~ Manpower equivalents Consultant Higher surgical trainee Staff grade Senior registrar Clinical assistant Basic surgical trainee Registrar Senior house officer
1.o 0.75
Details of morbidity and mortality came from 121 surgeons who returned data on deaths and/or complications covering 136 758 admissions and 119 588 operations (Table 6 ) . More than 100 surgeons gave overall complication rates and 85 details of types of complications. Only 41 could break these down into major and minor complications. Coloreetal resection A total of 2769 colonic resections were reported by 86 consultants (range 1 101 each). The mean mortality rate was 6 (range 0-43) per cent, the mean complication rate was 28 (range 0-211) per cent and the mean anastomotic leak rate 2.3 (range 0-123) per cent. The commonest complications were related to the abdominal wound (mean 9.8 (range 0-37) percent).Themean hospitalstay was 13.1 (range 2.3-27)days.
0.5
~
Service equivalent value equals the total value
Table 3 Interniediate procedures equioalents BUPA category Minor Intermediate Major Major plus Complex major A-D (now 1-5)
Intermediate equivalent value
Table 5 Proportion of patients treated in the lowe,~tand highest age groups
0.5
1.o 1.75 2.2 4.0*
*All complex major procedures were grouped together since there were so few in each category. The value reflects the fact that all were either C or D. BUPA, British United Provident Association
Patient age
No. of consultants
No. reported
Mean ( % )
< 10 years >60 years
122 122
7037 53 707
5 (0-13) 39 (17-67)
Values in parentheses are ranges
A meeting was held on 6 June 1991 at which the results were presented to surgeons who had taken part in the exercise. Each of these was given a questionnaire to assess reaction to the analysis and process.
Results Confidential numbers were requested by 220 (21 per cent) of the surgeons contacted. Of these, 160 (73 per cent) returned data on some or all of the items requested. Surgeons returned data when available; it was made clear that it was not necessary to complete the whole pro forma to participate and, therefore, variable numbers appeared in each of the analyses. The results of the workload analyses are shown in Table 4 . The discrepancy in variable junior staff is corrected by dividing the total number of admissions per year by the service equivalent value. The complexity of operations performed is corrected by dividing the total by the intermediate equivalents value. Some of the results of the case mix treated are shown in Table 5 and Figure 2 . Surgeons were given personal figures for each diagnosis (Figure 2 ) and for operative groups treated.
a
a
J
0
Figure 2 Diagnostic case mix. Data were requested for each of ten diagnostic groups; the rest of the admissions fell into other miscellaneous groups
Table 4 Resources and workload ~~
No. of consultants Resources Manpower available (SEV) Available theatre sessions per week Cancelled theatre sessions per year Beds available Occupied beds per day
111 110 67 100
Workload Overall admissions Overall operations Day-case admissions Non-operative admissions Emergency admissions ( YO) Admissions per SEV Intermediate equivalent value per year Intermediate equivalent value per SEV Mean stay length
127 122 80 114 126 111 97 88
No. reported
1.58 (0.5-2.75) 3 (1-7) 14 (0-55) 19 (5-40) 12 (1-29)
105
105
Mean
147 882 122 620 26 384 30 529 47 980 84 232 92241
1164 (121-2216) 1005 (121-2071) 329 (20-937) 268(2-950) 34 (7-71 ) 759 (183-1673) 951 (134-2271 ) 626 (195-1211) 4 (1-8)
Values in parentheses are ranges. SEV, service equivalent value
1074
Br. J. Surg.. Vol. 79, No. 10, October 1992
Audit of general surgical admissions: D. C. Dunn and S. Fowler Table 6 Overall complications per admission
Mean rate per consultant consultants complications ( X ) Overall mortality Mortality after operation Overall complication rate Complications Wound Respiratory Haemorrhage Cardiovascular Gastrointestinal Genitourinary Metabolic Nervous system Miscellaneous Complication rate after operation Major complications after operation Minor complications after operation
No. of
No. of
121 121
2539 1183
2 (0-5) l(0-5)
100
6101
6 (0-22)
85 85 85 85 85 85 85 85 85 48
1306 1029 227 858 613 780 32 1 185 857 2795
1.4 (0-5.8) 1.1 (0-4'2) 0.2 (0-1.1) 0.9 (0-4.2) 0.7 (0-4.0) 0.9 (0-3.7) 0.3 (0-2.4) 0.2 (0-1.3) 1.1 (0-4.1) 6 (1-23)
41
929
2.7 (0-8.1 )
41
1106
3.2 (0-12.2)
Values in parentheses are ranges
Table 7 Results qf'suraey after the meeting of surgeons contributing to the audit
No. of
Question
positive responses
Was the meeting useful? Did the data surprise y o u ? Have you performed further analysis? Have you changed your practice? Should this exercise be repeated?
32 9 13 5 33
A total of 33 surgeons responded
Cholecystectomy Results were returned by 126 surgeons on 4885 cholecystectomies: 3212 open (111 surgeons) and 1673 laparoscopic (47 surgeons). Of the latter, 92 (5.5 per cent) were converted to open operation. Sixteen deaths were reported after open operation (0.50 per cent) and two (0.12 per cent) after laparoscopic surgery. The morbidity rates after these two procedures were 13.9 and 7.9 per cent respectively; the morbidity rate after conversion was 9.8 per cent. The mean hospital stay after open cholecystectomy was 7.4 (range 4-16) days and after the laparoscopic operation 2.2 (range 1-5 )days. Responses to survey Forty surgeons who supplied data and attended the meeting on 6 June 1991 were given a questionnaire and 33 responses were received. Table 7 shows that by 2 months after the analysis of the data 39 per cent of those who returned the survey had examined the data further and 15 per cent had altered their practice as a result.
Discussion Comparisons between the activities of different centres in the UK have been found to be of value in the past although these have often involved only studies of m ~ r t a l i t y ~ . ~ . With the advent of computerized surgical audit the opportunity arose for users of similar software to compare their annual results. This development was encouraged by The Royal College of Surgeons of England and the users of one such software system" held a meeting at the College in January 1990, followed by similar meetings". Based on the success of
Br. J. Surg.. Vol. 79, No. 10, October 1992
these meetings the College established a comparative confidential audit service. The benefits of setting up such a service are that surgeons are able to compare personal results with a large database, the size of which tends to compensate for irregularities in the collection of the data. Meaningful comparisons can therefore be made. Discordant results should stimulate an examination of local practice in an attempt to explain the differences. Usually these will be explained by local variations in practice but may highlight deficiencies in surgical technique or administration. This audit service was confidential and voluntary. Raw audit data are open to misinterpretation. For example, a hospital that has a policy of admitting patients with terminal cancer may be reflecting a caring attitude but will show a high mortality rate in audit studies. Insisting that such patients are treated at home will lower mortality rates but may not be in the interests of the local community or the patients. Such interpretations may be quoted in the popular press". If surgeons are encouraged to collect and analyse data, as described in this study, it may be possible to produce some standardized mortality rates that are of value to the public. Confidentiality removes the value of producing good results simply to enhance the status of a unit, by case selection, changing discharge policies and by fabricating results. In a confidential system there is no personal benefit from these manoeuvres. This service was designed to be voluntary. Compulsion would be a good way to ensure that a large amount of inaccurate data were included but surgeons who take part voluntarily are more likely to be interested in the result and therefore in the quality of data input. In setting up a comparative audit service a number of compromises must be made. A complicated pro forma will be a disincentive, yet if it is too simple no useful results will be obtained. By allowing surgeons to make returns excluding some items, a larger number were encouraged to take part. Those who did take part but were unable to produce mortality and morbidity figures were stimulated to develop such systems in the future. Many were unable to participate because of lack of data systems. If the authorities encourage and recommend audit then facilities must be provided. The overall complication and mortality rates in this and two previous studies at the College are similar'0.' '. The results of the laparoscopic cholecystectomy survey are close to those of the EuropeanI3 and North American14 studies. These similarities encourage the view that crude audit data are not wildly inaccurate in representing current surgical practice.
References 1.
2.
Dunn DC, Dale RF. Combined computer generated discharge documents and surgical audit. BMJ 1986; 292: 816-18. Ellis BW, Michie HR, Esufali ST, Pyper RJD, Dudley HAF. Development of a microcomputer based system for surgical audit and patient administration: a review. J R Sor Med 1987; 80: 157-61.
3. 4.
5. 6.
7. 8. 9. 10.
Dunn DC.Audit of a surgical firm by microcomputer: five years' experience. BMJ 1988; 2%: 687-91. Royal College of Surgeons of England. Guidelines for Surgicul Audit by Coniputer. London: RCSE, May 1991. Collins CD. Recommended values for use in surgical audit and surgical workload analysis. Ann R Coll Surg Engl 1991; 73(Suppl): 73-94. British United Provident Association. BUPA Srliedule o/ Procedures. London: BUPA, 1989. Jones SM, Collins CD. Caseload or workload? Scoring complexity of operative procedures as a means of analysing workload. BMJ 1990; 301: 324-5. Gilmore OJA, Griffiths NJ, Connolly JC et al. Surgical audit: comparison of the workload and results of two hospitals in the same district. BMJ 1980; 281: 1050-2. Nixon SJ. Regional Audit. Presentation at the Spring Clinical and Scientific meeting of the Royal College of Surgeons of Edinburgh. 1988. Dunn DC, Dale RF, Gumpert JRW. DuKy TJ. Combined surgical audit by microcomputer involving units in four health regions. Ann R Coll Surg Engl 1992; 74: 47-53.
1075
Audit 11.
12. 13.
of general surgical admissions: D. C. Dunn and S.Fowler
Emberton M, Rivett R, Ellis B. Comparative audit: a new method
with laparoscopiccholecystectomy. Ant JSury 1991: 161: 385-7. The Southern Surgeons Club. A prospective analysis of 1518 laparoscopiccholecystectomies. N E n y l J M e d 1991;324: 1073-8.
of delivering audit. Ann R Coll Sury Eriyl 1991; 73(Suppl):
14.
117-20. Toynbee P. Nervous surgeons’ best kept secret: their rate of success. Dai/j. Mail 1991: 19 November. Cuschieri A, Dubois F, Mouiel J et cil. The European experience
Paper accepted 25 June 1992
Short note Br. J. Surg. 1992, Vol. 79, October, 1076
Small bowel transit time in patients w i t h intra-abdominal adhesions
2200 20
I
0
a
D. M. Scott-Coombes, M. N. Vipond* and J. N. Thompson
0.
a
Departments of Surgery, Ealing Hospital, London and *George Eliot Hospital, Nuneaton, UK Correspondence to: Mr D. M. Scott-Coombes, Department of Surgery, Ealing Hospital, Uxbridge Road,
-
5
120
n
rn v)
80
’.
Patients and methods Fifteen patients were selected who had undergone operation within the past 3 years for extensive adhesional disease without significant ( > 20 cm) resection of small bowel and with no history of gastrectomy or vagotomy. The operations preceding the development of adhesional disease were appendicectomy (seven patients ). hysterectomy (three), oversewing of perforated peptic ulcer (two), caesarean section (two) and cholecystectomy (one). Another 15 patients. matched for age and sex. who had never suffered peritonitis or undergone laparotomy, were selected as controls. None of the 30 patients had evidence of any other gastrointestinal disease, and none was receiving any medication affecting bowel transit time. None suffered from any serious concomitant disease. All the patients with adhesions were asymptomatic at the time of analysis. The study was granted approval by the hospital ethical committee, and all 30 patients gave informed consent. The orocaecal transit time was measured using the lactulose hydrogen breath test. After an overnight fast, the patient ingested a standard lactulose meal (40 g glucose, 15 g Casilan (Crookes Healthcare, Nottingham, UK), 18 g corn oil, made up to 270 ml with water to which was added 30 g lactulose)3 and provided end expiratory breath samples at intervals of 10min. The breath hydrogen concentration was measured using a dedicated electrochemical cell (EC60; Bedfont Technical Instruments, Sittingbourne, Kent, UK). The orocaecal transit time was recorded as the time after ingestion at which there was a sustained rise in expired breath hydrogen 10 p.p.m. above the fasting level3.
-
-
ioa -
E
Southall, Middlesex UB1 3 E U , U K Intra-abdominal adhesions occur in over 90 per cent of patients following laparotomy Because adhesions cause considerable problems to both patients and surgeons’, much research has centred on methods to prevent their formation. The success of these methods is difficult to assess because repeat laparotomy is uncommon, and thus the only clearly defined clinical endpoint is the long-term incidence of adhesional bowel obstruction. The aim of this study was to measure small bowel transit time in patients with intra-abdominal adhesions and thus to determine the possible role of this technique in identifying patients with small bowel adhesions.
a
-
-
m ---ye
om omma
60
Figure 1 Orocuecul trunsit tinir in purirnrs wirh udhrsions und in controls. BUYSare niediari rulurs
Discussion The lactulose hydrogen breath test provides a simple non-invasive estimate of small bowel transit time4 and is considered to be a useful research tool for this purpose5. Intrasubject variability is reduced by the administration of lactulose in the form of a semiliquid meal’. The results of this study suggest that transit in patients with intra-abdominal adhesions may be delayed in the absence of symptoms or clinical evidence of obstruction. The observed overlap in transit time between the two groups shows that the test is not specific for intra-abdominal adhesions. Follow-up studies will be required to determine whether those patients with a more marked prolongation of transit time will prove to be at greater risk of developing clinical symptoms. In addition, the lactulose hydrogen breath test may prove useful in the assessment of methods designed to prevent formation of intra-abdominal adhesions.
References 1.
2. 3. 4.
Results
Menzies D, Ellis H. Intestinal obstruction from adhesions - how big is the problem? Ann R CON Surg D i g / 1990; 72: 60-3. McEntee G , Pender D, Mulvin D et a/. Current spectrum of intestinal obstruction. Br J Sury 1987; 74: 976-80. La Brooy SJ, Male P-J, Beavis AK, Misiewicz JJ. Assessment of the reproducibility of the lactulose H, breath test as a measure of mouth to caecum transit time. Gut 1983; 24: 893-6. Bond JH, Levitt MD. Investigation of small bowel transit time in man utilizing pulmonary hydrogen measurements. J Lab CIQi Med 1975; 85: 546-55. Gilmore IT. Orocaecal transit time in health and disease. Gut 1990; 31: 250-1.
Two control patients and one with adhesions did not produce hydrogen. The orocaecal transit time was prolonged in the adhesion group compared with the control group (median 140 versus 110 min, P < 0.05; Mann-Whitney U test) (Figure I ).
5.
1076
0007-1323/92/10107&01 0 1992 Butterworth-Heinernann Ltd
Paper accepted 10 May 1992
Comparative audit: an experimental study of 147 882
general surgical admissions during
I990
D.C. Dunn and S. Fowler Confidential Comparative Audit Service, The Royal College of Surgeons of England, 35-43 Lincoln’s Inn Fields, London WC2A 3PN, UK Correspondence to: Mr D. C. Dunn
In 1991,1025 general surgical Fellows of The Royal College of Surgeons of England were circulated with a pro forma and asked to submit local audit results f o r admissions during 1990 to a confidential comparative audit service. The individual topics of cholecystectomy and colorectal resection were studied. Data returned by 160 surgeons concerned 147882 admissions including 122 620 operations. Overall mortality rates ranged f r o m 0 to 5 per cent and morbidity rates f r o m 0 to 22 per cent. Laparoscopic cholecystectomy was associated with one-quarter of the mortality rate and two-thirds the morbidity rate of open cholecystectomy. Of the 33 surgeons who responded to a survey after the presentation of results, all wished to continue the exercise in future years; 39 per cent had been stimulated to perform further analyses and 15 per cent had changed practice habits as a result. Comparative audit involving large numbers of patients and surgeons is feasible and seems benejicial to participants.
Surgeons in the UK are required to audit practice. The ability to d o this varies and is often dependent on the availability of suitable information technology’.’. When information is available an understanding of local practice may be reached3. Results may then need to be compared within a larger group. This stimulus produced comparative audit and provided the opportunity to define what is required of information systems4. Comparative audit deals with ‘raw’ data, the collection of which is largely uncontrolled. Unevenness in data collection is, however, compensated for by the large numbers of patients involved. Information from such databases can then be used for comparison with unit or personal results. If investigation indicates that poor practice is the cause of difference, corrective action can be instituted. Comparison of individual working conditions can be made, giving a greater insight into the special characteristics of local practice. A confidential comparative audit service was set up at The Royal College of Surgeons of England and a pilot meeting organized for general surgeons to develop, test and evaluate the methodology.
Materials and methods
and surgical workload analysis were those a g r x d by a working party’ and are shown in Table 2. Surgeons graded their operations, where possible, into minor, intermediate and major according to the British United Provident Association classification6 and this grading was converted into the intermediate procedure equivalent value’ (Tuble 3). This enabled a weighted value to be produced for operative workload.
Table 1 Data requatcd
it7
tilt pro formu srtii to cotisultat~t.s
Resources Manpower. beds, theatre sessions Workload Admissions, emergencies, procedures, stay nights Case mix Age groups, diagnoses, operative groups Complications Mortality, complication grades and types Topics Cholecystectomy Numbers, deaths, complications Colorectal resection Numbers, deaths, complications
A booklet pro forma was designed and circulated to all general surgical Fellows of The Royal College of Surgeons of England working in England and Wales. Information was requested on activity in 1990; a summary of the items requested is shown in Table I . The pro forma wascirculated to 1025 surgeons at the end of March 1991 and completed returns requested by the end of April 1991. This time scale meant that only those with access to the data requested were likely to take part. Assurance was given that the data would be treated confidentially. A confidential number was allocated and all further handling of the data performed in connection with this number only. Some members of the secretariat had access to the names of the surgeons but no surgeon was given this information. Participants were asked to enter ‘n.a.’ where data were not available and ‘0’ for a nil return. Returned data were entered into a spreadsheet (Super Calc 5; Computer Associates, San Jose, California, USA). Each data item was ranked and means and 25.50 and 75 percentiles produced. Sheets were produced for each confidential number and the rankings and positions on each of the forty analyses could be seen. An example chart is shown in Figure I . Analyses were made of the manpower available to consultants. Recommended manpower equivalent values for use in surgical audit
0007-1323/92/101073~ 0 1992 Butterworth-Heinemann Ltd
1073
Audit of general surgical admissions: D. C. Dunn and S. Fowler Table 2 Recommended d u e s for use in surgical audit and surgical workload a n a l l s i . ~ ~ Manpower equivalents Consultant Higher surgical trainee Staff grade Senior registrar Clinical assistant Basic surgical trainee Registrar Senior house officer
1.o 0.75
Details of morbidity and mortality came from 121 surgeons who returned data on deaths and/or complications covering 136 758 admissions and 119 588 operations (Table 6 ) . More than 100 surgeons gave overall complication rates and 85 details of types of complications. Only 41 could break these down into major and minor complications. Coloreetal resection A total of 2769 colonic resections were reported by 86 consultants (range 1 101 each). The mean mortality rate was 6 (range 0-43) per cent, the mean complication rate was 28 (range 0-211) per cent and the mean anastomotic leak rate 2.3 (range 0-123) per cent. The commonest complications were related to the abdominal wound (mean 9.8 (range 0-37) percent).Themean hospitalstay was 13.1 (range 2.3-27)days.
0.5
~
Service equivalent value equals the total value
Table 3 Interniediate procedures equioalents BUPA category Minor Intermediate Major Major plus Complex major A-D (now 1-5)
Intermediate equivalent value
Table 5 Proportion of patients treated in the lowe,~tand highest age groups
0.5
1.o 1.75 2.2 4.0*
*All complex major procedures were grouped together since there were so few in each category. The value reflects the fact that all were either C or D. BUPA, British United Provident Association
Patient age
No. of consultants
No. reported
Mean ( % )
< 10 years >60 years
122 122
7037 53 707
5 (0-13) 39 (17-67)
Values in parentheses are ranges
A meeting was held on 6 June 1991 at which the results were presented to surgeons who had taken part in the exercise. Each of these was given a questionnaire to assess reaction to the analysis and process.
Results Confidential numbers were requested by 220 (21 per cent) of the surgeons contacted. Of these, 160 (73 per cent) returned data on some or all of the items requested. Surgeons returned data when available; it was made clear that it was not necessary to complete the whole pro forma to participate and, therefore, variable numbers appeared in each of the analyses. The results of the workload analyses are shown in Table 4 . The discrepancy in variable junior staff is corrected by dividing the total number of admissions per year by the service equivalent value. The complexity of operations performed is corrected by dividing the total by the intermediate equivalents value. Some of the results of the case mix treated are shown in Table 5 and Figure 2 . Surgeons were given personal figures for each diagnosis (Figure 2 ) and for operative groups treated.
a
a
J
0
Figure 2 Diagnostic case mix. Data were requested for each of ten diagnostic groups; the rest of the admissions fell into other miscellaneous groups
Table 4 Resources and workload ~~
No. of consultants Resources Manpower available (SEV) Available theatre sessions per week Cancelled theatre sessions per year Beds available Occupied beds per day
111 110 67 100
Workload Overall admissions Overall operations Day-case admissions Non-operative admissions Emergency admissions ( YO) Admissions per SEV Intermediate equivalent value per year Intermediate equivalent value per SEV Mean stay length
127 122 80 114 126 111 97 88
No. reported
1.58 (0.5-2.75) 3 (1-7) 14 (0-55) 19 (5-40) 12 (1-29)
105
105
Mean
147 882 122 620 26 384 30 529 47 980 84 232 92241
1164 (121-2216) 1005 (121-2071) 329 (20-937) 268(2-950) 34 (7-71 ) 759 (183-1673) 951 (134-2271 ) 626 (195-1211) 4 (1-8)
Values in parentheses are ranges. SEV, service equivalent value
1074
Br. J. Surg.. Vol. 79, No. 10, October 1992
Audit of general surgical admissions: D. C. Dunn and S. Fowler Table 6 Overall complications per admission
Mean rate per consultant consultants complications ( X ) Overall mortality Mortality after operation Overall complication rate Complications Wound Respiratory Haemorrhage Cardiovascular Gastrointestinal Genitourinary Metabolic Nervous system Miscellaneous Complication rate after operation Major complications after operation Minor complications after operation
No. of
No. of
121 121
2539 1183
2 (0-5) l(0-5)
100
6101
6 (0-22)
85 85 85 85 85 85 85 85 85 48
1306 1029 227 858 613 780 32 1 185 857 2795
1.4 (0-5.8) 1.1 (0-4'2) 0.2 (0-1.1) 0.9 (0-4.2) 0.7 (0-4.0) 0.9 (0-3.7) 0.3 (0-2.4) 0.2 (0-1.3) 1.1 (0-4.1) 6 (1-23)
41
929
2.7 (0-8.1 )
41
1106
3.2 (0-12.2)
Values in parentheses are ranges
Table 7 Results qf'suraey after the meeting of surgeons contributing to the audit
No. of
Question
positive responses
Was the meeting useful? Did the data surprise y o u ? Have you performed further analysis? Have you changed your practice? Should this exercise be repeated?
32 9 13 5 33
A total of 33 surgeons responded
Cholecystectomy Results were returned by 126 surgeons on 4885 cholecystectomies: 3212 open (111 surgeons) and 1673 laparoscopic (47 surgeons). Of the latter, 92 (5.5 per cent) were converted to open operation. Sixteen deaths were reported after open operation (0.50 per cent) and two (0.12 per cent) after laparoscopic surgery. The morbidity rates after these two procedures were 13.9 and 7.9 per cent respectively; the morbidity rate after conversion was 9.8 per cent. The mean hospital stay after open cholecystectomy was 7.4 (range 4-16) days and after the laparoscopic operation 2.2 (range 1-5 )days. Responses to survey Forty surgeons who supplied data and attended the meeting on 6 June 1991 were given a questionnaire and 33 responses were received. Table 7 shows that by 2 months after the analysis of the data 39 per cent of those who returned the survey had examined the data further and 15 per cent had altered their practice as a result.
Discussion Comparisons between the activities of different centres in the UK have been found to be of value in the past although these have often involved only studies of m ~ r t a l i t y ~ . ~ . With the advent of computerized surgical audit the opportunity arose for users of similar software to compare their annual results. This development was encouraged by The Royal College of Surgeons of England and the users of one such software system" held a meeting at the College in January 1990, followed by similar meetings". Based on the success of
Br. J. Surg.. Vol. 79, No. 10, October 1992
these meetings the College established a comparative confidential audit service. The benefits of setting up such a service are that surgeons are able to compare personal results with a large database, the size of which tends to compensate for irregularities in the collection of the data. Meaningful comparisons can therefore be made. Discordant results should stimulate an examination of local practice in an attempt to explain the differences. Usually these will be explained by local variations in practice but may highlight deficiencies in surgical technique or administration. This audit service was confidential and voluntary. Raw audit data are open to misinterpretation. For example, a hospital that has a policy of admitting patients with terminal cancer may be reflecting a caring attitude but will show a high mortality rate in audit studies. Insisting that such patients are treated at home will lower mortality rates but may not be in the interests of the local community or the patients. Such interpretations may be quoted in the popular press". If surgeons are encouraged to collect and analyse data, as described in this study, it may be possible to produce some standardized mortality rates that are of value to the public. Confidentiality removes the value of producing good results simply to enhance the status of a unit, by case selection, changing discharge policies and by fabricating results. In a confidential system there is no personal benefit from these manoeuvres. This service was designed to be voluntary. Compulsion would be a good way to ensure that a large amount of inaccurate data were included but surgeons who take part voluntarily are more likely to be interested in the result and therefore in the quality of data input. In setting up a comparative audit service a number of compromises must be made. A complicated pro forma will be a disincentive, yet if it is too simple no useful results will be obtained. By allowing surgeons to make returns excluding some items, a larger number were encouraged to take part. Those who did take part but were unable to produce mortality and morbidity figures were stimulated to develop such systems in the future. Many were unable to participate because of lack of data systems. If the authorities encourage and recommend audit then facilities must be provided. The overall complication and mortality rates in this and two previous studies at the College are similar'0.' '. The results of the laparoscopic cholecystectomy survey are close to those of the EuropeanI3 and North American14 studies. These similarities encourage the view that crude audit data are not wildly inaccurate in representing current surgical practice.
References 1.
2.
Dunn DC, Dale RF. Combined computer generated discharge documents and surgical audit. BMJ 1986; 292: 816-18. Ellis BW, Michie HR, Esufali ST, Pyper RJD, Dudley HAF. Development of a microcomputer based system for surgical audit and patient administration: a review. J R Sor Med 1987; 80: 157-61.
3. 4.
5. 6.
7. 8. 9. 10.
Dunn DC.Audit of a surgical firm by microcomputer: five years' experience. BMJ 1988; 2%: 687-91. Royal College of Surgeons of England. Guidelines for Surgicul Audit by Coniputer. London: RCSE, May 1991. Collins CD. Recommended values for use in surgical audit and surgical workload analysis. Ann R Coll Surg Engl 1991; 73(Suppl): 73-94. British United Provident Association. BUPA Srliedule o/ Procedures. London: BUPA, 1989. Jones SM, Collins CD. Caseload or workload? Scoring complexity of operative procedures as a means of analysing workload. BMJ 1990; 301: 324-5. Gilmore OJA, Griffiths NJ, Connolly JC et al. Surgical audit: comparison of the workload and results of two hospitals in the same district. BMJ 1980; 281: 1050-2. Nixon SJ. Regional Audit. Presentation at the Spring Clinical and Scientific meeting of the Royal College of Surgeons of Edinburgh. 1988. Dunn DC, Dale RF, Gumpert JRW. DuKy TJ. Combined surgical audit by microcomputer involving units in four health regions. Ann R Coll Surg Engl 1992; 74: 47-53.
1075
Audit 11.
12. 13.
of general surgical admissions: D. C. Dunn and S.Fowler
Emberton M, Rivett R, Ellis B. Comparative audit: a new method
with laparoscopiccholecystectomy. Ant JSury 1991: 161: 385-7. The Southern Surgeons Club. A prospective analysis of 1518 laparoscopiccholecystectomies. N E n y l J M e d 1991;324: 1073-8.
of delivering audit. Ann R Coll Sury Eriyl 1991; 73(Suppl):
14.
117-20. Toynbee P. Nervous surgeons’ best kept secret: their rate of success. Dai/j. Mail 1991: 19 November. Cuschieri A, Dubois F, Mouiel J et cil. The European experience
Paper accepted 25 June 1992
Short note Br. J. Surg. 1992, Vol. 79, October, 1076
Small bowel transit time in patients w i t h intra-abdominal adhesions
2200 20
I
0
a
D. M. Scott-Coombes, M. N. Vipond* and J. N. Thompson
0.
a
Departments of Surgery, Ealing Hospital, London and *George Eliot Hospital, Nuneaton, UK Correspondence to: Mr D. M. Scott-Coombes, Department of Surgery, Ealing Hospital, Uxbridge Road,
-
5
120
n
rn v)
80
’.
Patients and methods Fifteen patients were selected who had undergone operation within the past 3 years for extensive adhesional disease without significant ( > 20 cm) resection of small bowel and with no history of gastrectomy or vagotomy. The operations preceding the development of adhesional disease were appendicectomy (seven patients ). hysterectomy (three), oversewing of perforated peptic ulcer (two), caesarean section (two) and cholecystectomy (one). Another 15 patients. matched for age and sex. who had never suffered peritonitis or undergone laparotomy, were selected as controls. None of the 30 patients had evidence of any other gastrointestinal disease, and none was receiving any medication affecting bowel transit time. None suffered from any serious concomitant disease. All the patients with adhesions were asymptomatic at the time of analysis. The study was granted approval by the hospital ethical committee, and all 30 patients gave informed consent. The orocaecal transit time was measured using the lactulose hydrogen breath test. After an overnight fast, the patient ingested a standard lactulose meal (40 g glucose, 15 g Casilan (Crookes Healthcare, Nottingham, UK), 18 g corn oil, made up to 270 ml with water to which was added 30 g lactulose)3 and provided end expiratory breath samples at intervals of 10min. The breath hydrogen concentration was measured using a dedicated electrochemical cell (EC60; Bedfont Technical Instruments, Sittingbourne, Kent, UK). The orocaecal transit time was recorded as the time after ingestion at which there was a sustained rise in expired breath hydrogen 10 p.p.m. above the fasting level3.
-
-
ioa -
E
Southall, Middlesex UB1 3 E U , U K Intra-abdominal adhesions occur in over 90 per cent of patients following laparotomy Because adhesions cause considerable problems to both patients and surgeons’, much research has centred on methods to prevent their formation. The success of these methods is difficult to assess because repeat laparotomy is uncommon, and thus the only clearly defined clinical endpoint is the long-term incidence of adhesional bowel obstruction. The aim of this study was to measure small bowel transit time in patients with intra-abdominal adhesions and thus to determine the possible role of this technique in identifying patients with small bowel adhesions.
a
-
-
m ---ye
om omma
60
Figure 1 Orocuecul trunsit tinir in purirnrs wirh udhrsions und in controls. BUYSare niediari rulurs
Discussion The lactulose hydrogen breath test provides a simple non-invasive estimate of small bowel transit time4 and is considered to be a useful research tool for this purpose5. Intrasubject variability is reduced by the administration of lactulose in the form of a semiliquid meal’. The results of this study suggest that transit in patients with intra-abdominal adhesions may be delayed in the absence of symptoms or clinical evidence of obstruction. The observed overlap in transit time between the two groups shows that the test is not specific for intra-abdominal adhesions. Follow-up studies will be required to determine whether those patients with a more marked prolongation of transit time will prove to be at greater risk of developing clinical symptoms. In addition, the lactulose hydrogen breath test may prove useful in the assessment of methods designed to prevent formation of intra-abdominal adhesions.
References 1.
2. 3. 4.
Results
Menzies D, Ellis H. Intestinal obstruction from adhesions - how big is the problem? Ann R CON Surg D i g / 1990; 72: 60-3. McEntee G , Pender D, Mulvin D et a/. Current spectrum of intestinal obstruction. Br J Sury 1987; 74: 976-80. La Brooy SJ, Male P-J, Beavis AK, Misiewicz JJ. Assessment of the reproducibility of the lactulose H, breath test as a measure of mouth to caecum transit time. Gut 1983; 24: 893-6. Bond JH, Levitt MD. Investigation of small bowel transit time in man utilizing pulmonary hydrogen measurements. J Lab CIQi Med 1975; 85: 546-55. Gilmore IT. Orocaecal transit time in health and disease. Gut 1990; 31: 250-1.
Two control patients and one with adhesions did not produce hydrogen. The orocaecal transit time was prolonged in the adhesion group compared with the control group (median 140 versus 110 min, P < 0.05; Mann-Whitney U test) (Figure I ).
5.
1076
0007-1323/92/10107&01 0 1992 Butterworth-Heinernann Ltd
Paper accepted 10 May 1992
