Br. J. Surg. 1991, Vol. 78, August, 964-967
T. Bates, S. R . Ebbs*, M. Harrison and R . P. A’Hernt Departments of Surgery, William Harvey Hospital, Ashford, *King‘s College School of Medicine and Dentistry, London and ?Royal Marsden Hospital, London, UK Correspondence to: Mr T. Bates, Department of
Surgery, William Harvey Hospital, Ashford, Kent TN24 OLZ, UK
Influence of cholecystectomy on symptoms A group of 292 consecutive patients underwent cholecystectomy f o r gallstones with presumed biliary pain over a 4-year period and all completed a self-assessment questionnaire before operation. Over the following 2 years 18 patients died but no others were lost to follow-up. The remaining 274 patients completed a further questionnaire 1 and 2 years after operation. Demographic characteristics and abdominal symptoms have been compared with an age- and sex-matched control group using the same questionnaire. Before operation symptoms of flatulent dvspepsia were f a r more frequent in patients with gallstones but operation markedly reduced these symptoms to an incidence which almost matched that of the control group. However, 1 year after cholecystectoiny 34 per cent of patients still suflered some abdominal pain and of 35 patients referred back to hospital f o r investigation none has been shown to have a retained bile duct stone at a minimum follow-up of 5 years. A multivariate analysis indicated that preoperativejatulence together with long duration of attacks of pain are risk .factors f o r postoperative dissatisfaction as judged by a linear analogue scale. However, both these factors are common and neither is a good discriminator of a poor outcome. The prediction of a poor symptomatic outcome after cholecystectomy f r o m preoperative symptoms or patient characteristics had only limited success and all patients should be warned of this risk.
Cholecystectomy is an accepted treatment for the complications of gallstones and for biliary pain. However, the relationship between gallstones and the spectrum of symptoms described as flatulent dyspepsia is uncertain’ and it is therefore difficult to advise patients with these symptoms what the outcome of surgery is likely to be. The symptoms of flatulent dyspepsia were described by Johnson’ as: repeated belching, full feeling after a normal sized meal, inability to finish a normal sized meal, the abdomen becomes blown out so that clothes have to be loosened, burning in the upper abdomen, burning discomfort in the chest (heartburn), bitter fluid regurgitation into the mouth, vomiting or nausea. We have conducted a cohort study of all patients undergoing cholecystectomy on one firm at a District General Hospital over a 5-year period. Comparisons have been made with data collected from an age- and sex-matched control group admitted to the same hospital. The aim of the study was to clarify the effects of cholecystectomy upon symptoms in order that patients might be better counselled preoperatively.
Patients and methods Between July 1980 and April 1985, 295 consecutive patients undergoing cholecystectomy on one firm of the William Harvey District General Hospital, Ashford, Kent were asked to complete a self-assessment questionnaire preoperatively. They completed the questionnaire again on the anniversary of their operation 1 and 2 years later. The criteria for cholecystectomy were gallstones with upper abdominal pain which was considered to be of biliary origin or the presence of complications. No patient had a cholecystectomy for asymptomatic gallstones or for flatulent dyspepsia alone and patients were warned that all their symptoms might not be relieved. The questionnaire recorded demographic characteristics and the type, duration and frequency of abdominal pain and the presence of gastrointestinal symptoms. The questionnaires recorded all nine symptoms described by Johnson’ but for the purpose ofevaluation the following five categories were analysed: flatulence (belching), reflux-type indigestion, distension.
nausea and vomiting. The postoperative questionnaire also included a linear analogue scale on which the subject recorded their perception of the success of cholecystectomy. From February 1984 to August 1987. 278 patients admitted for elective surgery at the same hospital were asked to act as controls and complete the same preoperative questionnaire. In order to be eligible. individuals must not have had a history of biliary disease o r a diagnosis of any condition likely to produce abdominal symptoms. Controls were matched by sex and for age by decade. Body Mass Index (BMI = weight in kg/height in metres’) was calculated for all gallstone and control subjects. A copy of the questionnaire is available on application to the authors. The questionnaire data were entered into a ‘Simple’ database on a Prime 2655 computer (Prime UK Ltd., London, UK). Statistical analysis was performed using B M D P version April 1985 ( B M D P Statistical Software, Los Angeles, California, USA).
Results No patient refused to co-operate in the study or join the control group but three patients were too ill to complete the preoperative questionnaire and they all died in the early postoperative period. No patient was lost to follow-up over the ’-year period but in the first year one patient died of postoperative haemorrhage, seven of malignancy and six of cardiac, vascular or respiratory disease. In the second year there were four further deaths from malignancy. Control and cholecystectomy subjects appeared well-matched for demographic features (Table I), except that the gallstone patients were slightly heavier and had a higher BMI than the control group. The gallstone patients were more likely to designate themselves as housewives and gave a family history of gallstones more frequently. The classical description of the gallstone patient is the fat, fair, fertile, flatulent, female in the forties or but only 22 per cent fell within this age-sex range and had been pregnant. Flatulence (belching), distension and reflux-type indigestion were approximately twice as common in patients with gallstones
~
964
0007-1 323/91/080964-04
~~~
& 1991 Butterworth-Heinemdnn Ltd
influence of cholecystectomy on symptoms: T. Bates et al. Table I
Characteristics of gallstone
Mean age (range) Malej’female Never married Pregnancies Height (cm) Weight (kg) Body Mass Index (weightj’height’) Social status Professional Intermediate Skilled Unskilled Housewifet Smoker Under 2 units alcohol daily Family history of gallstones Oral contraceptive use
and control groups
Control n=278
Gallstone* n = 278
Significance1
52.8(21-95) (s.e. 0.94) 821196 22 2,27(0.12) 161 (0.6) 66.3(0.73) 23.4(2.14)
53.3(2@90) (s.e. 0.94) 821196 22 2.5q0.13) 164 (1.0) 67.3(0.78) 24.5(2.65)
ns. P=O,OI P = 0.28 P=O.oOI
6 24 116 8 92 182 224
4 19 90 10 126 178 222
P=O.OI ns. n.s.
45
70
P=O.Ol
77
70
n.s.
without these symptoms. However, the sex of the patient, the number or severity of previous attacks judged by patient scoring. previous hospital admissions or the need for exploration of the common bile duct did not affect the symptomatic outcome. Multivariate analysis was performed using stepwise logistic regression (Table 3). Preoperative flatulence and long versus short duration of attacks of pain appeared to predict a poor outcome and patients over 55 seemed to fare worse than younger patients. Persistence of abdominal pain postoperatively was associated with a poor symptomatic outcome and this also applied to recurrence of flatulence, nausea, fever and reflux-type indigestion. A total of 35 patients were referred back to hospital by their general practitioners and underwent investigation of their persisting postoperative symptoms. These records were reviewed in October 1990 at a minimum of 5 years
I
5c
*The gallstone group consists of all cholecystectomy patients who survived to complete a self-assessment questionnaire 1 year later; t Statistical analysis refers to females only, housewife versus others; 1r test or xz test with Yates continuity correction, one degree of freedom, used; n.s.. not significant
W 0,
m c
c
W
2 0,
a
25
1
Table 2
Sj~tiiptotiisrecordetiby gallstone patietiis couipured with control
gro1ip
Gallstone patients Control subjects ri=278
Pre-op n=278
~~~~~
Flatulence Distension Indigestion Nausea Vomiting Fever Rigor Abdominal pain Consulted general practitioner for pain
~
30
33 38 15 10 4 4 Nil Nil
~
66 57 71 62 54 34 30 100 100
0
2nd year post-op n=274
44 37 43 23 9 5 5 34 23
42 40 42 19 10 4
Figure 1
27
Table 3 Factors,found to be signiJirantly associated with cholerystectoniy being considered less than a complete success: multivariate analysis
..
-1
18
1
3
4
5
6
I
I
No better
Much improved
7
0
I
Complete success
The perrepion o / succes.~of c h o I ~ ~ c ~ ~ s t e c . i0 o n i1st ~ ~ .ycar; ~
. , 2nd yeur
Characteristic
than in controls (Table 2 ) . Nausea and vomiting-wertifour to five times as frequent but postoperatively the frequency of symptoms fell to near the levels reported by controls and there was little difference between the groups at 1 and 2 year follow-up. At 1 year, one-third of patients still had abdominal pain and one-quarter had consulted their general practitioner for this symptom. The linear analogue scale recordings of the perceived success of the operation (Figure I ) show that the cholecystectomy patients had very similar results at both one and 2 years postoperatively. From these results it was arbitrarily decided that a patient scoring less than nine on the linear analogue scale considered their operation less than a total success. A univariate analysis was performed on the results obtained from the questionnaires to determine which characteristics or symptoms were associated with the operation being considered less than a total success. Patients with preoperative flatulence (belching) ( P < 0.01), distension after meals, reflux-type indigestion or an alcohol intake of less than 2 units (P
T. Bates, S. R . Ebbs*, M. Harrison and R . P. A’Hernt Departments of Surgery, William Harvey Hospital, Ashford, *King‘s College School of Medicine and Dentistry, London and ?Royal Marsden Hospital, London, UK Correspondence to: Mr T. Bates, Department of
Surgery, William Harvey Hospital, Ashford, Kent TN24 OLZ, UK
Influence of cholecystectomy on symptoms A group of 292 consecutive patients underwent cholecystectomy f o r gallstones with presumed biliary pain over a 4-year period and all completed a self-assessment questionnaire before operation. Over the following 2 years 18 patients died but no others were lost to follow-up. The remaining 274 patients completed a further questionnaire 1 and 2 years after operation. Demographic characteristics and abdominal symptoms have been compared with an age- and sex-matched control group using the same questionnaire. Before operation symptoms of flatulent dvspepsia were f a r more frequent in patients with gallstones but operation markedly reduced these symptoms to an incidence which almost matched that of the control group. However, 1 year after cholecystectoiny 34 per cent of patients still suflered some abdominal pain and of 35 patients referred back to hospital f o r investigation none has been shown to have a retained bile duct stone at a minimum follow-up of 5 years. A multivariate analysis indicated that preoperativejatulence together with long duration of attacks of pain are risk .factors f o r postoperative dissatisfaction as judged by a linear analogue scale. However, both these factors are common and neither is a good discriminator of a poor outcome. The prediction of a poor symptomatic outcome after cholecystectomy f r o m preoperative symptoms or patient characteristics had only limited success and all patients should be warned of this risk.
Cholecystectomy is an accepted treatment for the complications of gallstones and for biliary pain. However, the relationship between gallstones and the spectrum of symptoms described as flatulent dyspepsia is uncertain’ and it is therefore difficult to advise patients with these symptoms what the outcome of surgery is likely to be. The symptoms of flatulent dyspepsia were described by Johnson’ as: repeated belching, full feeling after a normal sized meal, inability to finish a normal sized meal, the abdomen becomes blown out so that clothes have to be loosened, burning in the upper abdomen, burning discomfort in the chest (heartburn), bitter fluid regurgitation into the mouth, vomiting or nausea. We have conducted a cohort study of all patients undergoing cholecystectomy on one firm at a District General Hospital over a 5-year period. Comparisons have been made with data collected from an age- and sex-matched control group admitted to the same hospital. The aim of the study was to clarify the effects of cholecystectomy upon symptoms in order that patients might be better counselled preoperatively.
Patients and methods Between July 1980 and April 1985, 295 consecutive patients undergoing cholecystectomy on one firm of the William Harvey District General Hospital, Ashford, Kent were asked to complete a self-assessment questionnaire preoperatively. They completed the questionnaire again on the anniversary of their operation 1 and 2 years later. The criteria for cholecystectomy were gallstones with upper abdominal pain which was considered to be of biliary origin or the presence of complications. No patient had a cholecystectomy for asymptomatic gallstones or for flatulent dyspepsia alone and patients were warned that all their symptoms might not be relieved. The questionnaire recorded demographic characteristics and the type, duration and frequency of abdominal pain and the presence of gastrointestinal symptoms. The questionnaires recorded all nine symptoms described by Johnson’ but for the purpose ofevaluation the following five categories were analysed: flatulence (belching), reflux-type indigestion, distension.
nausea and vomiting. The postoperative questionnaire also included a linear analogue scale on which the subject recorded their perception of the success of cholecystectomy. From February 1984 to August 1987. 278 patients admitted for elective surgery at the same hospital were asked to act as controls and complete the same preoperative questionnaire. In order to be eligible. individuals must not have had a history of biliary disease o r a diagnosis of any condition likely to produce abdominal symptoms. Controls were matched by sex and for age by decade. Body Mass Index (BMI = weight in kg/height in metres’) was calculated for all gallstone and control subjects. A copy of the questionnaire is available on application to the authors. The questionnaire data were entered into a ‘Simple’ database on a Prime 2655 computer (Prime UK Ltd., London, UK). Statistical analysis was performed using B M D P version April 1985 ( B M D P Statistical Software, Los Angeles, California, USA).
Results No patient refused to co-operate in the study or join the control group but three patients were too ill to complete the preoperative questionnaire and they all died in the early postoperative period. No patient was lost to follow-up over the ’-year period but in the first year one patient died of postoperative haemorrhage, seven of malignancy and six of cardiac, vascular or respiratory disease. In the second year there were four further deaths from malignancy. Control and cholecystectomy subjects appeared well-matched for demographic features (Table I), except that the gallstone patients were slightly heavier and had a higher BMI than the control group. The gallstone patients were more likely to designate themselves as housewives and gave a family history of gallstones more frequently. The classical description of the gallstone patient is the fat, fair, fertile, flatulent, female in the forties or but only 22 per cent fell within this age-sex range and had been pregnant. Flatulence (belching), distension and reflux-type indigestion were approximately twice as common in patients with gallstones
~
964
0007-1 323/91/080964-04
~~~
& 1991 Butterworth-Heinemdnn Ltd
influence of cholecystectomy on symptoms: T. Bates et al. Table I
Characteristics of gallstone
Mean age (range) Malej’female Never married Pregnancies Height (cm) Weight (kg) Body Mass Index (weightj’height’) Social status Professional Intermediate Skilled Unskilled Housewifet Smoker Under 2 units alcohol daily Family history of gallstones Oral contraceptive use
and control groups
Control n=278
Gallstone* n = 278
Significance1
52.8(21-95) (s.e. 0.94) 821196 22 2,27(0.12) 161 (0.6) 66.3(0.73) 23.4(2.14)
53.3(2@90) (s.e. 0.94) 821196 22 2.5q0.13) 164 (1.0) 67.3(0.78) 24.5(2.65)
ns. P=O,OI P = 0.28 P=O.oOI
6 24 116 8 92 182 224
4 19 90 10 126 178 222
P=O.OI ns. n.s.
45
70
P=O.Ol
77
70
n.s.
without these symptoms. However, the sex of the patient, the number or severity of previous attacks judged by patient scoring. previous hospital admissions or the need for exploration of the common bile duct did not affect the symptomatic outcome. Multivariate analysis was performed using stepwise logistic regression (Table 3). Preoperative flatulence and long versus short duration of attacks of pain appeared to predict a poor outcome and patients over 55 seemed to fare worse than younger patients. Persistence of abdominal pain postoperatively was associated with a poor symptomatic outcome and this also applied to recurrence of flatulence, nausea, fever and reflux-type indigestion. A total of 35 patients were referred back to hospital by their general practitioners and underwent investigation of their persisting postoperative symptoms. These records were reviewed in October 1990 at a minimum of 5 years
I
5c
*The gallstone group consists of all cholecystectomy patients who survived to complete a self-assessment questionnaire 1 year later; t Statistical analysis refers to females only, housewife versus others; 1r test or xz test with Yates continuity correction, one degree of freedom, used; n.s.. not significant
W 0,
m c
c
W
2 0,
a
25
1
Table 2
Sj~tiiptotiisrecordetiby gallstone patietiis couipured with control
gro1ip
Gallstone patients Control subjects ri=278
Pre-op n=278
~~~~~
Flatulence Distension Indigestion Nausea Vomiting Fever Rigor Abdominal pain Consulted general practitioner for pain
~
30
33 38 15 10 4 4 Nil Nil
~
66 57 71 62 54 34 30 100 100
0
2nd year post-op n=274
44 37 43 23 9 5 5 34 23
42 40 42 19 10 4
Figure 1
27
Table 3 Factors,found to be signiJirantly associated with cholerystectoniy being considered less than a complete success: multivariate analysis
..
-1
18
1
3
4
5
6
I
I
No better
Much improved
7
0
I
Complete success
The perrepion o / succes.~of c h o I ~ ~ c ~ ~ s t e c . i0 o n i1st ~ ~ .ycar; ~
. , 2nd yeur
Characteristic
than in controls (Table 2 ) . Nausea and vomiting-wertifour to five times as frequent but postoperatively the frequency of symptoms fell to near the levels reported by controls and there was little difference between the groups at 1 and 2 year follow-up. At 1 year, one-third of patients still had abdominal pain and one-quarter had consulted their general practitioner for this symptom. The linear analogue scale recordings of the perceived success of the operation (Figure I ) show that the cholecystectomy patients had very similar results at both one and 2 years postoperatively. From these results it was arbitrarily decided that a patient scoring less than nine on the linear analogue scale considered their operation less than a total success. A univariate analysis was performed on the results obtained from the questionnaires to determine which characteristics or symptoms were associated with the operation being considered less than a total success. Patients with preoperative flatulence (belching) ( P < 0.01), distension after meals, reflux-type indigestion or an alcohol intake of less than 2 units (P
