A longitudinal study of risk factors associated with the formation of pressure ulcers in nursing homes

Abstract

Objective: To determine risk factors associated with the formation of stage II-IV pressure ulcers in nursing homes.

Design: Since the incidence rate for pressure ulcer formation varies among nursing homes, the homes were divided into tertiles based on these rates. Pooled logistic regression was used to model which factors are associated with the formation of pressure ulcers in both high and low incidence homes.

Setting: 78 National HealthCorp nursing homes.

Subjects: We studied 4232 nursing home residents free of pressure ulcers on admission to a nursing facility and at 3-months follow-up. All remained in the home for at least 3 additional months to a maximum of 21 months.

Measurements: The effects of age, gender, race, antipsychotic drug use, urinary incontinence, fecal incontinence, body mass index, diabetes mellitus, disorientation, ambulation, physical restraints, activities of daily living of bathing, feeding, or transferring, and nursing home bed size on the formation of a stage II-IV pressure ulcer while the subject was a resident in the nursing home were studied.

Main results: Significant factors associated with the formation of pressure ulcers in high incidence homes (21-month incidence = 19.3%) were ambulation difficulty (OR = 3.3; CI = 2.0, 5.3), fecal incontinence (OR = 2.5; CI = 1.6, 4.0), diabetes mellitus (OR = 1.7; CI = 1.2, 2.5), and difficulty feeding oneself (OR = 2.2; CI = 1.5, 3.3). In the low incidence homes (21-month incidence = 6.5%), significant factors associated with pressure ulcer incidence were ambulation difficulty (OR = 3.6; CI = 1.7, 7.4), difficulty feeding oneself (OR = 3.5; CI = 2.0, 6.3), and male gender (OR = 1.9; CI = 1.2, 3.6).

Conclusions: Although low and high incidence homes share similar risk factors, such as ambulation and feeding activities of daily living, the main difference was that diabetes and fecal incontinence played a major role only in high risk homes, while male gender was an important discriminator only in low incidence homes. Yet, it is unclear if these factors explain the three-fold difference in the incidence rates for pressure ulcers in these facilities. Baseline or resident clinical characteristic differences of any one factor between the high and low incidence homes varied by no more than 5%. While we identified certain conditions which are associated with pressure ulcer formation, there may be unknown or unmeasured facility effects in addition to the characteristics of a given resident in a particular home.