Background HIV services face the challenge of regular monitoring of a growing patient cohort and ensuring prompt action upon abnormal results. We identified a number of issues using a paper-based results system (PBS) including: (1) missing results, (2) delayed delivery, (3) clinician error, (4) lack of audit trail, placing patients at risk of delayed identification of drug toxicity and serious conditions for example, acute hepatitis.
Aims We piloted an electronic results checking system (ERC) which classifies results as normal or abnormal (non-urgent (NUAbn) or urgent (UAbn)) to compare the speed and performance of PBS and ERC in identifying biochemical abnormalities.
Methods Between 4 July 2011 and 22 July 2011 we compared the time intervals from sampling to (1) receipt of results; (2) clinician review of UAbn/NUAbn and (3) review of NUAbn by a clinician. Abnormalities were graded and both systems reviewed daily. Data were analysed using STATA V.11.0. Mann–Whitney U tests were used to compare the intervals.
Results Of 513 patients undergoing ≥ one blood test, 296 (57.7%) had ≥ one biochemical abnormality identified by the ERC (10.7 % UAbn, 42.3% NUAbn and 47% not clinically significant). Of these, PBS simultaneously identified 83%. The median interval between sampling to (1) receipt of results was 1 (IQR 1–2) vs 4 days (IQR 3–5), p<0.0001; (2) clinician review 3 (IQR 1–4) vs 3 (IQR 3–6) days, p<0.037; and (3) review of NUAbn by clinician 2 (IQR 1–4) vs 10 days (IQR 9–12), P=0.136, for ERC and PBS respectively. 11% of the missing PBS results were classified UAbn. ERC missed three abnormalities highlighting a software error which has now been corrected.
Conclusion We demonstrate the use of IT to review blood results leads to the faster identification of biochemical abnormalities, which are common in our HIV cohort, facilitating their timely management. We anticipate the use of ERC in routine practice will avoid delay/non-identification of a significant number of abnormal results within our service.