Risk Prediction Model for 6-Month Mortality for Patients Discharged to Skilled Nursing Facilities.

The study was retrospective and used administrative and electronic record data, so it could not fully describe how care was delivered. It came from one integrated health system in the US Midwest with a largely White patient population, so it may not apply elsewhere. Health care use and some outcomes outside the system may have been missed. The analysis did not include socioeconomic status, nutritional status, medications, or cardiopulmonary resuscitation preferences. It could not confirm whether patients had short-term or long-term skilled nursing facility stays. Facility quality, care processes, length of stay, and health care use during the skilled nursing facility stay were not considered. The model was checked with internal cross-validation and should be confirmed in other health care systems before routine use. Each discharge was treated as an independent encounter, even when the same patient had more than one discharge.

The authors declare no conflicts of interest. The article is listed as having research support from the National Institutes of Health, Extramural.

Do not treat the model as ready for every skilled nursing facility. It was developed and checked only in one mostly White US Midwest health system, and the authors state that results should be confirmed in other health care systems.