The impact of territory-wide surveillance on postoperative hypothermia among elective surgical patients in a major public hospital in Hong Kong

Abstract

Introduction In the perioperative period, patients are more prone to hypothermia as a results of excessive heat loss due to an impaired temperature homeostasis. Unintentional perioperative hypothermia is associated with a number of adverse post-operative outcomes. A variety of measures for prevention of unintentional perioperative hypothermia is available for different patients of different needs. Clinical practice varies between departments and physicians. Gaps exist between knowledge from best available evidence and research, and what happens in clinical practice. Performance measurement and evaluation is one of the means to continuously improve the health care quality and safety.

A Temperature Monitoring Project has been implemented on all anaesthesia departments in Hospital Authority (HA) hospitals to quantify the occurrence of unintentional perioperative hypothermia. The data collected were summarised every three months by each anaesthesia department and sent to the secretary of the Central Coordinating Committee for further compiling and analysis. This information is subsequently fed back to all anaesthesia departments for further action. Each department would then implement their own measures to reduce the incidence of unintentional perioperative hypothermia.

This study was conducted with patient data from Tuen Mun Hospital, one of the major general public hospitals in Hong Kong. Given that the surveillance system is a territory-wide intervention across all public hospitals in Hong Kong, using a similar tertiary-wide surveillance system on other clinically important outcomes may further enhance the quality and safety of our health care delivery and hence improving the health outcomes of our population.

Methods The aim of this study was to determine whether implementing a territory-wide surveillance on post-operative hypothermia by the COC has an impact on outcome among patients who received elective non-cardiac surgical procedures in Tuen Mun Hospital. This was an uncontrolled before and after study at two time points. Patients older than 10 years of age who had elective non-cardiac surgical procedures performed during the first period in January to March 2015 were compare with the patients from the second period in January to March 2016. Patient data used in this study was retrieved through multiple electronic databases including Anaesthesia Computerised Information System (ACIS), Clinical Data and Analysis Report System (CDARS) and Electronic Patient Records (ePR). Datasheets collected through different electronic databases were matched by unique Hospital Numbers (HN) of the patients.

Univariate comparisons were performed for the demographics, the characteristics of anaesthesia and operation, and the occurrence of hypothermia. Further univariate analysis was performed to identify the potential associations between the occurrence of hypothermia and the patient demographics and characteristics of anaesthesia and operation. Logistic regression was used to estimate the odds of the occurrence of hypothermia between the two study periods. Subgroup analysis was performed to determine the relative changes of occurrence of hypothermia between the two periods of implementation of surveillance system between patients receiving operations from different surgical teams.

Results There was a total of 1112 and 1095 eligible patients receiving elective non-cardiac operation in Tuen Mun Hospital from the first and second study period respectively. There is no statistical significant difference in the demographic distribution of the patients, and the characteristics of anaesthesia and surgical procedure between the two study periods. Implementation of the territory-wide surveillance of post-operative hypothermia has reduced the incidence of hypothermia in elective non-cardiac operations from 18.5% to 12.7% which is also statistically significant (χ2 = 19.78, p <0.001). Logistic regression for occurrence of hypothermia in 2016 compared with 2015 shows that the implementation of the Temperature Monitoring Project reduces the chance of developing post-operative hypothermia by 36.5% (Odds ratio 0.635). The patients undergoing surgical procedures by surgery, orthopaedic surgery and thoracic surgery were found to be associated with a higher risk of developing postoperative hypothermia.

Conclusion Unintentional perioperative hypothermia is still an important issue that we need to address in order to improve the quality and safety of our patients in the perioperative period. The results from this study shows an association between the implementation of territory-wide surveillance program for post-operative hypothermia and reduction of its occurrence in our patients, hence potentially improve the health care quality and safety. Further investigation is required to determine whether the reduction in occurrence in post-operative hypothermia reduce hypothermia related perioperative complications and which interventions implemented as a result of the territory-wide surveillance has contributed to this effect. Questions still remain on whether the impact of surveillance program can be sustained and whether it can be applied in other areas of clinical medicine.