VAHI’s Health Data Integrity (HDI) program aims to ensure that health data collections accurately reflect the care provided to patients and coding to clinical truth. The Health Data Integrity Program plan 2018–19 to 2019–20 published last year, supports the transition of the HDI Program from an emphasis on random reviews (or audits) designed to achieve statewide estimates of overall accuracy, to a more targeted approach based on data analytics and risk assessment.
Accordingly, VAHI has taken a more targeted hospital and patient record approach in year two of the Victorian Admitted Episodes Dataset (VAED) audits and concentrated on the accuracy of hospital acquired complications (HACs) data derived from diagnoses and prefix coding. A HAC is defined as a serious complication that a patient has suffered after admission to hospital. VAHI reports information about HACs in its statewide quarterly Inspire reports.
VAHI selected 45 health services for audits during 2018–19. This includes metropolitan, regional and small rural health services that report acute and, where applicable, mental health data.
In each of the chosen health services, a stratified random sample method is used to select the acute care episodes to be reviewed in the audit. The sample design is used to provide insights into the accuracy of reporting of HACs as well as to assess the reliability of the data used to measure the actual incidence of HACs.
Preliminary results from audits at the first 22 health services indicate that the overall level of reported HACs in the audit sample is ‘about right’. The audit sample included 275 episodes where the health service reported a HAC as compared to 288 episodes identified by the audit to have a HAC.
However, this overall result masks variations in the accuracy of HACs’ reporting.
Among the 275 episodes where a HAC had been reported by a health service, the auditors found 81 episodes that did not meet the criteria for coding as a HAC, either due to incorrect prefix assigned (37) or the diagnosis code not confirmed by the audit (44). The auditors also found an additional 94 episodes in the sample where a HAC should have been reported.
The auditors also identified a lot of variation across and within the health services, with 12 health services under-reporting HACs overall and eight health services over-reporting the incidence of HACs.
The audits are progressing with the remaining 23 health services. The focus on HACs will continue into year three of the audit program, with one of the main priorities for VAHI to improve HACs coding in partnership with health services.
For more information, please contact Beata Steinberg, Acting Manager, Health Data Integrity, firstname.lastname@example.org