Coordinated care affects hospitalization and prognosis in amyotrophic lateral sclerosis: a cohort study.

BACKGROUND: To determine whether an integrated approach to coordination of care influences hospitalization and clinical outcomes in a chronic neurological disease, amyotrophic lateral sclerosis. METHODS: We followed up 2452 patients with probable or definite amyotrophic lateral sclerosis from 2000 to 2012. Two cohorts were compared before and after the creation of a community care network for this disease in Ile de France in 2006. During these two periods, the medical and paramedical care teams and formal standards of care were identical; the only difference was the coordination by the network. To investigate hospital and emergency department use, we used number of patients, number of stays, and number of days. For clinical outcomes, we used slopes of functional deterioration, and Kaplan-Meier and Cox models for survival. RESULTS: All hospitalization variables decreased after the creation of the network, which was not explained by admissions elsewhere. The slope of functional deterioration was significantly different before (1.03 ± 1.57 points/month) and after (0.79 ± 0.80 points/month; p = 0.002) creation of the network. Patients included in the network had a median survival time of 13.2 months more (log rank test; p < 0.001). In the Cox model, the network intervention was associated with a 45% decrease in relative risk of death during the period of the study (p < 0.001). CONCLUSIONS: Network care was associated with fewer hospital admissions, reduced functional deterioration and later mortality in ALS. These results suggest that proactive coordination between carers in chronic and complex diseases could have a positive impact on hospitalization and the clinical course of the disease.

Outcomes in older patients after surgical treatment for hip fracture: a new approach to characterise the link between readmissions and the surgical stay.

BACKGROUND: in older patients, there is a high risk of hospital readmission within the first year after surgery for hip fracture, due to complications following treatment or to the evolution of prior comorbid conditions. OBJECTIVES: to identify factors associated with readmissions related to the index surgical stay. DESIGN: retrospective cohort study. SETTING: administrative claims databases. SUBJECTS: patients over 75 surgically treated for hip fracture in Paris area. METHODS: we analysed all admissions in 2005, and tracked for 1-year readmissions. First readmissions (FRs) were classified as related or unrelated to the index stay, according to rules defined a priori. We analysed the association between patient characteristics and the FR. RESULTS: among 5,709 patients, 32% had at least one readmission, 53% were FR related. Near 80% of related readmissions occurred within 3 months from discharge. Surgical conditions caused 47% of all related readmissions, and male gender, dementia, cancer or kidney diseases were independent risks factors. CONCLUSIONS: half of readmissions could be classified as related to the index stay and a great majority of these occurred early post discharge. Surgical conditions caused 47% of all related readmissions. Improvement in orthopedic-geriatric co-care is suitable to expect an impact on outcomes after surgery.

Were less disabled patients the most affected by 2003 heat wave in nursing homes in Paris, France?

OBJECTIVE: To analyse the change of mortality rates (MRs) and their contributing medical factors among nursing home patients during the 2003 heat wave in France. METHODS: A retrospective observational study was conduced in all nursing homes of the Assistance-Publique-Hôpitaux de Paris (AP-HP), the French largest public hospital group. All AP-HP nursing home patients (4,403) who were institutionalized in May, 2003, were concerned. The MRs of patients between three periods (before, during and after the August 2003 heat wave period) were compared according to their demographic characteristics, level of dependence and medical condition. RESULTS: The MR increased from 2.2 per cent persons month (ppm) (1.9-2.4) before heat wave up to 9.2 ppm (8.0-10.4) during heat wave and back to 2.4 ppm (2.2-2.7) after heat wave. MRs before heat wave were higher among highly dependent patients compared to those less dependent [mortality rate ratio (MRR) = 2.66 (1.69-4.21)]. This difference disappeared during the heat wave [MRR = 1.28 (0.91-1.81)] and appeared again after heat wave [MRR = 2.21 (1.52-3.23)]. The same pattern was observed for several medical conditions, such as severe malnutrition or swallowing disorders. CONCLUSION: These results suggest that medical care during heat wave has been directed towards more fragile patients, helping to limit deaths in this group. Less frail patients made the largest contribution to excess mortality during the heat wave. During extreme weather conditions, specific attention should be paid not only to frail persons, but to all the elderly community.

Measuring performance in health care: case-mix adjustment by boosted decision trees.

OBJECTIVE: The purpose of this paper is to investigate the suitability of boosted decision trees for the case-mix adjustment involved in comparing the performance of various health care entities. METHODS: First, we present logistic regression, decision trees, and boosted decision trees in a unified framework. Second, we study in detail their application for two common performance indicators, the mortality rate in intensive care and the rate of potentially avoidable hospital readmissions. RESULTS: For both examples the technique of boosting decision trees outperformed standard prognostic models, in particular linear logistic regression models, with regard to predictive power. On the other hand, boosting decision trees was computationally demanding and the resulting models were rather complex and needed additional tools for interpretation. CONCLUSION: Boosting decision trees represents a powerful tool for case-mix adjustment in health care performance measurement. Depending on the specific priorities set in each context, the gain in predictive power might compensate for the inconvenience in the use of boosted decision trees.

A regression shrinkage method tailored to qualitative regressors and clustered data.

We propose a shrinkage method for estimation in linear regression models with qualitative regressors. Due to the nature of the shrinkage constraint, this method tends to give estimates that are exactly zero for some groups of coefficients belonging to the same regressor. The method hence results in concise models, since some of the regressors are entirely eliminated. In conjunction with this estimation method, a model with a fixed cluster effect turns out to be closely related to frailty models. We apply the method for modelling hospital readmissions.