Using linked hospitalisation data to detect nursing sensitive outcomes: a retrospective cohort study

Area of Science:

• Healthcare research
• Nursing science
• Health informatics

Background:

• Nursing-sensitive outcomes are adverse patient events linked to nursing care.
• Current methods for identifying these outcomes using administrative data face limitations in adjusting for pre-existing conditions.
• Incorporating patient history through look-back periods can enhance accuracy.

Purpose of the Study:

• To evaluate the impact of linked hospitalisation data and extended look-back periods on the incidence rates of nursing-sensitive outcomes.
• To investigate associations between patient demographics, hospitalisation characteristics, and nursing-sensitive outcomes.

Main Methods:

• Retrospective cohort study conducted at three tertiary metropolitan hospitals in Western Australia.
• Analysis of linked hospitalisation data over a 5-year period.
• Application of a 2-year look-back window to identify prior relevant health information.

Main Results:

• A reduction in estimated incidence rates for all nursing-sensitive outcomes was observed when a 2-year look-back period was applied.
• Survival analysis indicated most relevant disease information is captured within 2 years.
• Patients with nursing-sensitive outcomes were significantly older, more likely female, had comorbidities, and longer stays.

Conclusions:

• Current methods may over-estimate nursing-sensitive outcome rates.
• Linked hospitalisation data with look-back periods improve the accuracy of detecting these outcomes.
• This approach offers a more precise method for identifying adverse events using administrative data.