OBJECTIVES

: Use of routine patient data has been strongly encouraged by public funders of healthcare research in the UK. Routine data are available from national sources e.g. Hospital Episode Statistics but there can be a delay in obtaining this data and it may be incomplete.

We introduced a method for obtaining resource use data directly from 26 hospitals participating in the UPSTREAM trial (ISRCTN56164274). This replaced Case Report Forms (CRFs) which are manually completed from medical notes and take up significant research nurse time. We present the estimated cost to hospitals.

METHODS

: We contacted participating hospital informatics teams and requested two data downloads; an interim download twelve months from trial completion and a “top-up” download on completion. Information analysts were invited to complete feedback questionnaires. These included the time taken to complete the work from coding to verification and the number of datasets queried.

RESULTS

: All 26 participating hospitals returned diagnostic (ICD-10), procedure (OPCS-4) and currency codes (HRGs) for inpatient stays, outpatient appointments and A&E attendances. Requests for feedback were sent with two follow-up reminders. 58% of analysts responded. For these centres, the average time taken for the initial data extraction was 187 minutes (95%CI: 121–241 minutes). Patients per centre ranged from 7–114 and number of datasets 2–7. NHS Information analysts and nurses follow the same salary structure, therefore any time saving manifests as a cost saving.

CONCLUSIONS

: It is possible to obtain a complete dataset of inpatient, outpatient and A&E resource use directly from hospitals. The potential for missing data is lower and the method is likely to be cost saving over nurse completed CRFs.

Funded by the National Institute for Health Research HTA programme. The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the HTA programme, NIHR, NHS or the Department of Health.