Draining the gallbladder before surgery added a week in hospital and doubled ICU admission. It did not reduce complications

When a gallbladder becomes acutely inflamed and the patient looks too fragile for an operation, the standard alternative is to put a drain in it. A needle through the liver into the gallbladder relieves the pressure, the fever settles, and surgery waits for a better day. It is a reasonable idea, it is done frequently in Korean hospitals, and two recent Korean studies have now measured what it costs.

Neither found the benefit the practice assumes.

What the guideline actually says#

The Tokyo Guidelines 2018, whose authorship includes several Korean hepatobiliary surgeons, set out a flowchart by severity grade. For grade I disease it recommends early laparoscopic cholecystectomy where the Charlson comorbidity index is 5 or less and the ASA physical status is 2 or less. For grade II it makes the same recommendation, adding that the operation should be performed by an experienced surgeon; only if those criteria are not met does it direct the patient to medical treatment and gallbladder drainage first. Grade III can still be operated on early, but under stricter conditions and at an advanced centre.

The guideline also dropped the older 72-hour window. Early now means within seven days of symptom onset rather than within three, a change that widened the group of patients eligible for immediate surgery.

The Hallym series: ten days instead of four#

A team led from Hallym University Dongtan Sacred Heart Hospital reviewed 505 patients who underwent laparoscopic cholecystectomy for grade II or III acute cholecystitis between 2012 and 2022, then matched 122 who had received drainage against 122 who went straight to theatre. The results were published in the American Journal of Surgery.

The authors' conclusion is unusually plain for a surgical paper: drainage may extend hospitalisation and intensive care use without improving outcomes.

The delay that predicted harm came before admission#

The same analysis tested which delays mattered. The interval that independently predicted postoperative complications was not the wait between admission and operation. It was the wait before the patient arrived: presenting more than four days after symptoms began tripled the odds of a complication, with an odds ratio of 3.01 (95 percent confidence interval 1.16 to 7.81, p = 0.023).

That finding points somewhere other than the operating theatre. It is a statement about how quickly people with upper abdominal pain reach a hospital that can operate on them, which is a question of triage and referral rather than of surgical technique.

Four centres, 835 patients, no difference in biliary complications#

The larger of the two studies, published in Annals of Surgical Treatment and Research, drew on 835 patients with grade II or III disease treated at four Korean tertiary centres. After one-to-one matching it compared 201 patients who had drainage first against 201 who had early laparoscopic cholecystectomy, and found no difference in biliary complications: a hazard ratio of 1.103, with a confidence interval from 0.519 to 2.343 that comfortably spans no effect.

Its authors read this as evidence that early laparoscopic cholecystectomy is feasible in most cases of moderate-to-severe cholecystitis, while keeping drainage for patients whose underlying disease makes general anaesthesia itself the hazard. That is a narrower indication than current practice, not an argument against drainage.

For patients who do get a drain, the interval is measurable#

A single-centre study at Konyang University Hospital followed 695 patients with grade I or II cholecystitis who had drainage and then surgery between 2010 and 2019, and looked for the intervals at which outcomes deteriorated.

• In grade I disease, more than 3.5 days from symptom onset to drainage predicted a technically difficult operation.
• In grade I disease, more than 7.5 days from drainage to cholecystectomy predicted a poor postoperative outcome.
• In grade II disease, neither interval was statistically related to surgical difficulty or postoperative outcome.

The grade I thresholds are derived from receiver operating characteristic analysis in one hospital's records and have not been validated prospectively elsewhere. They are a starting point for scheduling, not a standard.

What these studies cannot show#

All three are retrospective, and they share one limitation that no amount of statistical adjustment removes.

1. Patients receive drains because they are judged too unwell for immediate surgery, so the drainage group is sicker before any treatment decision is recorded. Propensity matching corrects only for the characteristics that were measured.
2. The four-centre study shows this directly: after matching, the drainage group still had significantly more preoperative systemic inflammatory response syndrome (24.9 versus 6.5 percent), more pneumonia (7.5 versus 3.0 percent) and more cardiac disease (67.2 versus 57.7 percent) than the surgical group.
3. Longer stays and more intensive care in the drainage group are therefore consistent with two explanations that these designs cannot separate: that the drain delayed recovery, or that the patients who received drains were always going to need more care.