Completed Therapy Worksheet for Evidence-Based General Surgery
Kiviluoto T, Siren J, Luukkonen P & Kivilaakso E. Randomised trial of laparoscopic versus open cholecystectomy for acute and gangrenous cholecystitis. Lancet 1998; 351: 321-5.
Are the results of this therapy study valid?
Was the assignment of patients to treatments randomised?
Were all patients who entered the trial accounted for at the conclusion? Were they analysed in the groups to which they were randomised?
Were patients and clinicians kept blind as to which treatment was being received?
Aside from the experimental treatment, were the groups treated equally?
The management of suspected common bile duct stones was different in the two groups: This may have prolonged the mean operating time of the patients undergoing open cholecystectomy, because 60% of them had an operative cholangiogram added to the procedure.
Were the groups similar at the start of the trial?
Are the valid results of this randomised study important?
End points considered: hospital morbidity, length of hospital stay, length of time off work, percentage of LC converted to OC.
It is worth pointing out that, whilst physicians are usually interested in the percentage benefit from their treatments, and the percentage of their patients in which this occurs, there are many situations in which surgeons can reasonably expect a 100% response rate in all of their patients (as here, where it is impossible to get cholecystitis after the operation). In these situations, comparisons of different techniques must necessarily concentrate on minimising the adverse effects of the surgery. Many surgical publications deal mainly with these effects: The commonly used, generally applicable measures of adverse effects are:
- Morbidity (defined as any adverse event related to the treatment, but usually applied only to COMPLICATIONS, defined as adverse effects of treatment which are NOT INEVITABLE)
- Hospital Stay
- Pain (from the operation)
- Time off Work
Morbidity and Mortality can be expressed as rates, and an NNT calculated. Durations can usefully be expressed as an index which is 1/duration. Hospital stay can then be compared using the ratio of indices e.g. Hospital Stay Index for LC (median) = 1/4 = 0.25. HSI for OC is 1/6 = 0.167. The Hospital Stay Ratio between the two is therefore 6/4 = 1.5. Another method (useful only if everyone can agree on a suitable figure) is to calculate the percentage of cases in which the duration involved exceeded a given figure, and the NNT to prevent one instance of this calculated. This is the procedure used here for duration of time off work.
|OC Complication Rate
|LC Complication Rate
|42%||3% (19%)*||93% (55%)*||38% (23%)*||2.6 (4.3)*|
95% CI on NNT = +/- 0.25 (0.21)
* (IF CONVERSION TO OC IS COUNTED AS A COMPLICATION)
Length of morbidity
|OC median stay
|LC median stay
|6 days||4 days||1.5|
Length of time off work
|OC sick leave > 2/52*
|LC sick leave > 2/52*
95% CI on NNT = +/- 0.06
* arbitrary cut-off point chosen by reviewer
Percentage of LC converted to OC
Calculations inapplicable: rate was 16%
Can you apply this valid, important evidence in caring for your patient?
Do these results apply to my patient?
Is your patient so different from those in the trial that its results can't help you?
How great would the potential benefit of therapy actually be for my patient?
Method I: Risk of (complications or conversion to OC) in my patient relative to risk in the trial, expressed as a decimal : presumed to be 1.
NNT/f = 4.3
Method II: Your patient's expected event rate if they received the control treatment:
1 / (PEERxRRR)
= 1 / (42%x55%)
Are my patient's values and preference satisfied by the regimen and its consequences?
Needs to be assessed in each patient, but YES for the patient described.
Clinical Bottom Line
This was a small study, and the difference in the rate of complications between the groups was really quite surprising, since most uncontrolled reports have shown the biggest risk of complications is amongst LC patients who undergo conversion to OC. The LC were all carried out by experts, whilst the OC were mostly carried out by trainees. As with many surgical trials, adequate blinding could not be carried out. This may have influenced length of hospital stay (see Squirrel et al, Surgery 1998; 123: 485-95). Overall, the conclusion that LC is better has to be regarded with some caution.