Endocrine Abstracts

Introduction: The National Paediatric Diabetes Audit (NPDA) provides a benchmark of performance for paediatric diabetic services across the UK. Whether intentional or not, a league table is created comparing units based on their mean HbA1c. Although the coordinators suggest submitting the patients’ most recent HbA1c, this may not necessarily be a universally adopted phenomenon. We examined the effect of selecting patient’s best, yearly average, and latest HbA1c on our unit’s overall mean HbA1c and its impact on our position in the ‘league’.

Design: All patient HbA1c values were collected for two NPDA periods, January 2010–March 2011 and January 2011–March 2012. From the four HbA1c results collected during that period, the patients best, average for the year and last HbA1c were compared. The impact of omitting ‘occasional’ higher results was also examined. Our clinic average HbA1c was calculated and compared to Yorkshire regional and National data to assess the impact on our ranking.

Results: For the period 2010–2011, our clinic mean HbA1c varied significantly from 8.0% using best HbA1c to 8.5% with average HBA1c, moving us from 2nd to 13th in the league table (historical data). There was a significant difference of 8.0 vs 8.4% taking best rather than most recent HbA1c values. Similarly, for 2011–2012, there were significant variations from 7.8 to 8.2 and 8.3% using the best, last and mean HbA1c variables.

Conclusion: Dependent on the variable submitted a clinically relevant difference of 0.5% was noted in the overall mean HbA1c. Such a difference could see you as champions or candidates for relegation in the league table! The system is potentially open to foul play and tighter regulation of selection policy is required as HbA1c is increasingly used as a performance indicator and in some cases the basis for quality payments.