Quality Improvement Cycles that Reduced Waiting Times at Tshwane District Hospital Emergency Department

Abstract

Background: Tshwane District Hospital (TDH) is a level-one hospital, delivering services in the centre of Pretoria since February 2006. It is unique in location, being only 100 meters away from the tertiary hospital, Pretoria Academic Hospital (PAH). In South Africa, public sector emergency units are under enormous pressure with large patient numbers, understaffing and poor resources. TDH Emergency Department (ED) is a typical example. An average of 3 900 patients per month visited this ED in 2006. Recurrent complaints and dissatisfaction shown by the patients about prolonged waiting times before consulting the medical practitioners (MPs) in the ED were one of the initial challenges faced by the newly established hospital. It was decided to undertake quality improvement (QI) cycles to analyse and improve the situation, using waiting time as a measure of improvement. Methods: A QI team was chosen to conduct two QI cycles. The allocated time for QI cycle 1 was from May to August 2006 and for QI cycle 2 from September to December 2006. A total of 150 waiting times of stable and unstable patients were evaluated. Fifty waiting times were recorded over a span of 24 hours for each data collection in May, September and December 2006. Waiting time was defined as the time from arrival of the patient in the unit until the start of the consultation by the MP. Surveys were done in May and September to analyse the problems causing prolonged waiting times. The implemented change included instituting a functional triage system, improvement of the process of up- and down-referrals to and from the tertiary hospital, easy access to stock, reorganisation of doctors' duty roster, reorganisation of the academic programme, announcement on waiting time to patients, nurses carrying out minor procedures and availability of reference books.
Results: The median waiting times for stable patients were as follows: May 2006: 545 minutes (range 200 to 1 260), September 2006: 230 minutes (range 15 to 480) and December 2006: 89 minutes (range 15 to 230). There was a significant difference
among these waiting times for May, September and December 2006 (p < 0.001; Kruskal-Wallis H test). The median waiting times for unstable patients were as follows: May 2006: zero minutes (range 0 to 30), September 2006: zero minutes (range 0 to 3)
and December 2006: 0.5 minutes (range 0 to 2). There was no difference among the waiting times for unstable patients for May, September and December 2006 (p = 0.90; Kruskal-Wallis H test). Conclusion: This QI exercise identified problems causing prolonged waiting time for stable patients at TDH ED. It rectified most of the identified problems. However, goals regarding registration and laboratory delays could not be successfully achieved. This study showed the significance of QI cycles in improving waiting times for stable patients at TDH ED without any additional financial or human resources. This was done without compromising the time taken to see unstable patients.

South African Family Practice Vol. 50 (6) 2008: pp. 43-43e