How to switch a manual record collection with an automatic calculation from medico- administrative databases for quality indicators? Example of the quality indicator " Care project, Life project " in rehabilitation care
Sophie Baron a, Marie-Caroline Clement a, Robin Louvel a, Pauline Renaud a, Joëlle Dubois a
Introduction
The hospital Financial Incentive for Quality Improvement program (IFAQ) has been launched in France in 2012 by the Ministry of Health.
Currently, quality indicators used for IFAQ were developed by the High Health Authority (HAS) and are based on data collected from patient files selected at random each year. This data collection represents an important workload for hospitals. So, to reduce it, the Ministry of Health asked Technical Agency for Information on Hospital Care (ATIH) to explore the feasibility to switch the current record with the hospital medico-administrative databases (PMSI) for the calculation.
The "Care Project, Life Project" indicator (PSPV) for rehabilitation care centers (SSR) has been selected. The aim was to analyze the possibility to calculate it from PMSI instead of patients' record with similar results.
Methods
HAS indicators are calculated through an assessment form with 8 criteria (physical examination, social or psychological evaluation) regarding the care and the life projects elaborated for the patient during his stay. At the end, a score is calculated per stay, then aggregated at the hospital level.
Through the PMSI SSR in France, rehabilitation acts (CSARR acts) and activity daily life (ADL) score are routinely collected for each stay and could be used to calculate this indicator. Thus, an algorithm has been developed to approach each criteria from PMSI SSR mainly using both, CSARR acts and ADL. It has then been tested to estimate PSPV for all the stays corresponding to the indicator perimeter used for the drawing lots from 2017 to 2019.
The algorithm results were first analyzed per criteria and per hospital; and were then compared to those obtained from HAS in 2018.
Results
6 on 8 criteria could be approached from PMSI SSR. The ADL score criteria was better for all hospitals from PMSI SSR than from patients' records. However, scores for the five other criteria were lower.
The PSPV global score per hospital was lower from the PMSI SSR than from PSPV records.
Conclusions
HAS approved the calculation based only on 6 criteria as missing criteria do not alter the validity of the indicator measurement.
Results shown poor agreement between the two calculation methods. This could be related to a problem with the CSARR coding (misunderstanding of the wording or lack of completeness).
To further this work, the PMSI draw tools have been improved for the 2021 patient record data collection. This will allow to compare results of the two methods on a stay-by-stay basis and to adapt, if necessary, the PMSI algorithm (Results expected in September 2022). Meanwhile, ATIH will launch campaign to raise hospitals' awareness of the need to improve the recording of CSARR acts by the PMSI. This improvement will be necessary for the changeover without losing the quality of the indicator.
Even if the first results of the PSPV from the PMSI are the worst, this method of calculating indicators from medico-administrative databases should be preferred in the future. Indeed, it avoids increasing the workload of hospitals and allows the results to be calculated on all the activity, and not just on a sample, which increases the robustness of the indicator.
a ATIH, France
Original Version in PDF