RANDOMIZATION TO SURGERY OR OBSERVATION IN PATIENTS WITH SYMPTOMATIC GALLBLADDER STONE DISEASE - THE PROBLEM OF EVIDENCE-BASED MEDICINE IN CLINICAL-PRACTICE

Background: The description and understanding of gallbladder stone dis ease in the medical literature are difficult because an assessment of the natural course of this symptomatic disease, with separation of pat ients in strictly defined groups, is rally lacking. Method: A multicen ter study was carried out with patients randomized to either surgery o r conservative, expectant treatment to examine optimal treatment and n atural history in well-defined groups of symptomatic gallbladder stone disease with pain episodes only (study group 1) or acute cholecystiti s (study group 2). The patients were between 18 and 80 years of age an d had right upper quadrant or midline epigastric pain and ultrasonogra phic evidence of gallbladder stone, with or without acute cholecystiti s. Medical treatment was ordinated on the basis of signs and symptom s everity. Patients randomized to surgery were placed on the hospital's waiting list and electively operated on with cholecystectomy as soon a s conveniently possible. Preliminary results of follow-up are based on questionnaires mailed al regular intervals and consultations if requi red by the patients' symptoms. Quality of life and freedom from pain a re study end points for evaluation. Results: Five hundred and one pati ents were considered for the study, 318 patients (63.5%) belonging to group 1 and 183 (36.5%) to group 2. One hundred and eighty-four patien ts (36.7%) were randomized, 120 (24%) in group 1 and 64 (12.7%) in gro up 2. Three hundred and seventeen patients (63.3%), with an equal prop ortion from each study group, were excluded, the largest number of pat ients because of incapacitating pain or serious presenting symptoms (2 6%). More women (81%) than men (19%) had pain alone (group 1), bur the sex ratio was reduced, 56% women and 44% men, in patients with acute cholecystitis (group 2): with a highly significant difference (P = < 0 .001) between the two groups. A significant difference in patient with drawal from randomized treatment was registered, with 24% from observa tion and 12% from surgery (P = 0.032), but with an equal distribution in the two study groups. Conclusions: A randomized trial of this natur e is feasible but extremely difficult to perform because of the hetero geneous nature of gallbladder stone disease, leading to exclusion of m any patients and difficulties in measuring and evaluating outcome vari ables.