PT - JOURNAL ARTICLE AU - Magda Bucholc AU - Maurice O’Kane AU - Brendan O’Hare AU - Ciaran Mullan AU - Paul Cavanagh AU - Siobhan Ashe TI - Effects of a demand optimization intervention on laboratory test utilization in primary care AID - 10.1101/573956 DP - 2019 Jan 01 TA - bioRxiv PG - 573956 4099 - http://biorxiv.org/content/early/2019/03/12/573956.short 4100 - http://biorxiv.org/content/early/2019/03/12/573956.full AB - There is evidence of increasing use of laboratory tests with substantial variation between clinical teams which is difficult to justify on clinical grounds. The aim of this project was to assess the effect of a demand optimisation intervention project on laboratory test requesting by general practitioners (GPs) in an area of Northern Ireland supported by the Clinical Chemistry Laboratory service of Western Health and Social Care Trust (WHSCT). The intervention package was developed in conjunction with the Western Local Commissioning Group and consisted of educational initiatives, feedback to 55 individual practices on test request rates with ranking relative to other practices, and a small financial incentive for practices to reflect on their test requesting activity. Overall test utilization rates of profile tests, HbA1c, and PSA one year before, during, and one year after the intervention were measured using laboratory databases of the Altnagelvin Area Hospital, Tyrone County Hospital, and the Erne (South West Acute Hospital. The intervention was associated with mixed effects. First, we observed a reduction of 5.1% in the median profile test request rates and a decrease in their between practice variability. The overall downward trend in variability of profile test request rates was found statistically significant (p = 0.03). Second, we found a significant increase in both the volume (p < 0.0001) and between practice variability (p = 0.0001) of HbA1c requests per patient with diabetes. The increase in HbA1c requests may reflect a more appropriate rate of diabetes monitoring and also the adoption of HbA1c as a diagnostic test. Yet, the subsequent 600% increase in between practice variability of HbA1c ordering rates may imply an inconsistent implementation of recommended guidelines by GPs. Finally, there was a 29.3% increase in the median and 35% increase in between practice variability of request rates for PSA, the reasons for which are unclear.