29 JUNE 2001 - NURSE STAFFING IN THE US ASSOCIATED WITH HOSPITAL PATIENT OUTCOMES

 

Hospital nurse staffing levels continue to be of concern to the United States Government as the median age of the US population grows older and nurse shortages become more chronic. Implicit in this concern is the belief that nurse staffing levels and the quality of hospital care are associated.

To investigate this link, the Healthcare Financing Administration, the Agency for Healthcare Research and Quality, and the National Institutes of Nursing Research recently sponsored a comprehensive retrospective study. Results from this study support the concern that nurse shortages may adversely impact the quality of care in the US.

Studying hospital patient discharge data

This study used hospital patient discharge data to examine the association between nurse staffing levels and what the study called “outcomes potentially sensitive to nursing” (OPSNs). Information on nurse staffing levels was derived from hospital staffing surveys and state hospital financial reports. The study used three separate samples:

· Sample 1 was made up of 799 hospitals from eleven states (Arizona, California, Massachusetts, Maryland, Missouri, Nevada, New York, South Carolina, Virginia, Wisconsin, and West Virginia). It was used to investigate nurse staffing at the hospital level and 14 OPSNs in all patients and in Medicare patients

· Sample 2 surveyed 256 California hospitals, which were used to examine inpatient staffing for ICU, acute and long-term patient services

· Sample 3 was a national sample of 3,357 hospitals, which were used to investigate the appropriateness of using Medicare hospital discharge data to measure the association of nurse staffing and patient outcomes (information was derived from the American Hospital Association annual survey of hospitals)

The study analyses investigating the association between nurse staffing and patient outcomes controlled for a number of factors likely to confound the relationship, including: hospital size, location, occupancy rate and teaching status. In addition, a patient's likelihood of developing one of the complications examined in the study was summarized in a risk measure derived from the patient's age, sex, diagnostic category, comorbid disease conditions, emergent admission and payment type.

Results from the nursing/patient study

Study results from Sample 1 indicate a strong association between nurse staffing and five patient outcomes in medical patients: urinary track infections, pneumonia, length of stay, upper GI bleeds and shock. Among major surgery patients, the only consistent relationship found was between nurse staffing and the “failure to rescue” (death among patients with sepsis, pneumonia, shock, deep vein thrombosis/pulmonary embolism or GI bleeding). Sample 2 was generally consistent with Sample 1. Analysis of Sample 3 indicates that it is acceptable to use Medicare data to assess patient outcomes in instances where data for all patients are not available.

The study authors recommend improved data collection to better capture information on secondary diagnoses, admission status, more accurate nurse staffing documentation. This includes updated procedures for calculating full- and part-time equivalent employees for RNs, LPNs and nursing aides. Such data collection improvements will facilitate future monitoring and system improvements.