Prospective cohort study to determine if perioperative mortality at the University Teaching Hospital in Lusaka, Zambia has changed compared to historical data from 1987

Abstract

Despite low-income and middle-income countries (LMIC) being the regions of the world with the greatest gap in access to safe surgical and anaesthetic services, there are fewer published data for these countries on perioperative mortality indices. Global trends are that perioperative mortality has declined over the last five decades. Zambia has had two published studies over the last 30 years. The initial study conducted by Heywood, Wilson and Sinclair in 1987 revealed an overall sixth day inpatient perioperative mortality rate (POMR) of 7.55 per 1000 operations which was more than three times the current rate of high income countries. Surgical avoidable POMR was highest at 1.93 per 1000 operations, followed by administrative avoidable POMR at 0.85 per 1000 operations and then anaesthetic avoidable POMR lowest at 0.53 per 1000 operations. Some of the recommendations from this study resulted in the establishment of a trainining programme in anaesthesia and critical care, and procurement of modern anaesthetic and monitoring equipment. It was the lack of more recent knowledge to ascertain whether the quality of the Zambian surgical and anaesthetic services have improved in keeping with global trends that made it imperative to conduct a perioperative mortality study in the largest teaching hospital in Zambia as a follow up to the initial study. Over a six months period we prospectively identified all patients who died within six days of surgery, having undergone a surgical procedure by either a general, regional or combined anaesthesia at the University Teaching Hospital in Lusaka. All demised patients had their records analysed by a specialist surgeon and anaesthetist to determine factors contributing to their death by consensus. The causes of death were classified as avoidable, partially avoidable and unavoidable. The deaths categorised as avoidable and partially avoidable were further categorised into contributing factors, namely anaesthetic, surgical, administrative cause or a combination thereof. The data on denomenator was collected from the surgical and recovery room registers. The numerator was the total number of all inpatient deaths within six days of the surgical procedure, with date of surgery being day one. A total of 9775 cases were captured with 449 of these being lost to follow up. The denominator was therefore made of 9326 cases. Seventy nine patients were detected to have died during the study period. The overall sixth day inpatient perioperative mortality rate (overall POMR) was found to be 0.85% (95% CI: 0.68-1.06) and the sixth day inpatient avoidable mortality rate (avoidable POMR) was found to be 0.42% (95% CI: 0.30-0.57). The sixth day inpatient anaesthetic POMR, surgical POMR and administrative POMR were found to be 0.04% (95% CI: 0.01-0.079), 0.19% (95% CI: 0.12-0.31) and 0.19% (95% CI: 0.12- 0.31) respectively. Compared to the historical indices from 1987 by Heywood, Wilson and Sinclair the chi-square test revealed a difference of no statistical significance (x2=0.411 with 1df, p=0.522). With regard to the 24 hour inpatient perioperative mortality indices, the overall POMR was found to be 0.30% (95% CI: 0.21-0.44). The 24hr inpatient avoidable mortality rate (avoidable POMR) was 0.17% (95% CI: 0.10-0.28). The 24hr inpatient anaesthetic, surgical and administrative avoidable POMR were 0.02% (95% CI: 0.00-0.08), 0.07% (95% CI 0.03-0.16) and 0.08% (95% CI: 0.03-0.16) respectively. The perioperative mortality indices at UTH have not reduced in comparison to the historical data from 30 years ago. This is contrary to global indices which have shown a downward trend over the last five decades. Keywords: perioperative, mortality, numerator, denominator, avoidable