Nine cohort studies with historical controls (at least 9,943 participants; numbers not reported in three studies) were included. Study quality was low to moderate. Conflict of interest was not reported in most studies, the study design was open to bias, methods of participant selection and allocation concealment and blinding were mostly unclear, data collection methods were only partly explained and some studies used partly appropriate analysis methods.
Intensive care units (five studies): All studies set in intensive care units used different types of checklists, reported different outcomes and were at high risk of bias. One study reported a significant decrease in length of stay for the checklist group and another reported that the percentage of ventilator days per week when patients received all four care processes increased significantly from 30% before to 96% during the intervention. In another study only four domains from the checklist were assessed and compared and this found significantly better results during the intervention period for use of physical therapy and transfer to telemetry. Two other studies reported reductions in length of stay but did not report whether they were statistically significant.
Emergency departments (two studies): Both studies set in emergency departments were at high risk of bias. One used a safety checklist for patients with an indwelling urinary tract catheter and found appropriate use of catheters increased following the intervention but this was not statistically significant. There was a significant increase in physician orders for catheter placement. In a study of a post-endoscopy checklist there was a significant decrease in length of stay with use of the checklist.
Surgery (one study): This study had a moderate risk of bias. It assessed a 19-item surgical safety checklist in eight hospitals in eight countries. Rates of major complications, deaths in hospital, surgical site infections and unplanned reoperations fell significantly during the intervention period. Process adherence to correct surgical protocols that comprised six safety measures showed that the intervention significantly improved five of the six measures. Only the outcome of ensuring the presence of at least two peripheral intravenous catheters or a central venous catheter before incision in cases with an estimated blood loss of 500mL or more did not improve.
Acute care (one study): This study had a high risk of bias. It found hospitals that used a checklist administered appropriate antibiotics within eight hours for patients with pneumonia significantly more often than hospitals without the checklist.