Is it legitimate to use unplanned hospitalizations as a quality indicator for cancer patients? A retrospective French cohort study with special attention to the influence of social deprivation.

PURPOSE: Readmission indicators are used around the world to assess the quality of hospital care. We aimed to assess the relevance of this type of indicator in oncology, especially for socially deprived patients. Our objectives were (1) to assess the proportion of unplanned hospitalizations (UHs) in cancer patients, (2) to assess the proportion of UHs that were avoidable, i.e., related to poor care quality, and (3) to analyze cancer patients the effect of patients' deprivation level on the type of UH (avoidable UHs vs. unavoidable UHs). METHODS: In a French university hospital, we selected all hospitalizations over a year for a random sample of cancer patients. Based on medical records, we identified those among UHs due to avoidable health problems. We assessed the association between social deprivation, home-to-hospital distance, or home-to-general practitioner with the type of UH (avoidable vs. unavoidable) via a multivariate binary logit estimation. RESULTS: Among 2349 hospitalizations (355 patients), there were 383 UHs (16 %), among which 38% were avoidable. Among UHs, the European Deprivation Index was significantly associated with the risk of avoidable UHs, with a lower risk of avoidable UH for patients with medium or high social deprivation. CONCLUSION: Our results suggest that the use of UHs rate as a quality indicator is questionable in oncology. Indeed, the majority of UHs were not avoidable. Furthermore, within UHs, those involving patients with medium or high social deprivation are more often unavoidable in comparison with other patients.

The impact of a patient's social status on the cost of vaginal deliveries: an observational study in a French university hospital.

INTRODUCTION: We aimed to assess the association between a patient's social status and the cost of stay for a single uncomplicated vaginal delivery. Currently, few data have been reported. MATERIAL AND METHODS: We conducted an observational study with data retrieved from the medical and administrative databases of a university hospital in North-West France. We included all patients admitted in 2014 and classified in either Diagnosis-Related Group (DRG) « Single uncomplicated vaginal deliveries in a primiparous patient ¼ or DRG « Single uncomplicated vaginal deliveries in a multiparous patient ¼. Criteria defining poor social status were: a specific healthcare benefit in relation to low income or for foreign undocumented patients, and/or a consultation with a social worker during the hospital stay except if no social problem was diagnosed. We compared the cost of stay between patients with poor social status and patients with good social status using a multivariate median regression stratified on parity, and adjusted for age, gestational age and neonatal hospitalization. RESULTS: Among 686 primiparous patients, 21% had poor social status, which was associated with an increase in the median cost of stay (+€475; 95% CI [+334 to +616]), mostly explained by a 1-day increase in the median length of stay.Among 899 multiparous patients, 29% had poor social status, which was not associated with the cost of stay. DISCUSSION: Social status had an impact on the cost of vaginal deliveries in primiparous patients. Our findings suggest a need to redefine the DRG classification according to patients' social status.