Exploring preventive interventions and risk factors of hospital-acquired pressure ulcers: A retrospective matched case-control design.
Previous literature showed weak and sometimes contradictory evidence regarding the best interventions to prevent pressure ulcers and the best factors that can serve as predictors for ulceration. The aim of this study was to explore effective interventions and associated risk factors in the area of pressure ulcer. A retrospective approach was used to explore such interventions and risk factors in a more natural clinical environment than found in a prospective study. While retrospective studies have their limitations, one problem of prospective studies, the Hawthorn effect, is not present. In order to meet the aims of the study, a matched case-controlled design was employed. A convenience sampling technique was used to select all patients who matched the study criteria. Two groups of patients were selected. The first group developed pressure ulcer during hospitalization, the other did not. In order to have a sound and robust comparison, each patient from the pressure ulcer groups was matched or at least nearly matched with another patient from the non-pressure ulcer group for a number Waterlow sub-scores. Further criteria for selection included a minimum of three days total length of stay in hospital and being initially free of any pressure ulcer on admission for both of the study groups. Electronic medical records for all patients were revised, and multidimensional data were extracted using a data extraction sheet. Data analyses were carried out using univariate analysis (t-test, Mann-Whitney, Chi-square and Fisher’s exact test) and multivariate analysis (binary logistic regression). In univariate analysis for preventive interventions, the following interventions were significantly associated with pressure ulcer prevention (P≤ 0.05): standard hospital bed, seating cushion, static pressure redistributing mattress, re-positioning every four hours and helping the patient to sit regularly in a chair. When the effect of all interventions was adjusted through the multivariate model, the following interventions were independently associated with prevention: draw sheet, re-positioning every four hours and helping patient to sit regularly in chair (odds ratio = 0.24, 0.06 and 0.13 respectively; P≤ 0.05). In univariate analysis for risk factors related to physical activity and mobility, the following factors were significantly associated with developing pressure ulcer (P≤ 0.05): moving in bed with help, the ability to take a bath only in bed, needing two helpers in performing activities of daily living and moving outside bed only by a hoist. When adjusting the effect of all variables related to physical activity and mobility through the multivariate model, only two factors were independently associated with developing pressure ulcer: moving in bed with help and the ability to take a bath only in bed (odds ratio = 7.69 and 3.67 respectively; P≤ 0.05). In univariate analysis for risk factors related to pressure ulcer intrinsic risk factors, the following factors were significantly associated with developing pressure ulcer (P≤ 0.05): presence of three underlying medical conditions, dehydration, depression, having a blood transfusion, serum albumin <32mg/dl, haemoglobin <130 g/l in males or <115 for females and systolic blood pressure <113 mmHg. When adjusting the effect of all variables related to intrinsic risk factors through the multivariate model, the following risk factors were independently associated with pressure ulcer: presence of two underlying medical conditions, presence of three underlying medical conditions, cognitive impairment, serum albumin <32mg/dl and haemoglobin <130 g/l in males or <115 for females (odds ratio = 13.3, 143, 4.3, 0.10 and 0.14 respectively; P≤ 0.05). Findings from this study suggest a number of interventions to be effective in PUs prevention, and a number of risk factors that can predict risk of PUs. Findings were based on statistical association between acquiring PUs and the independent variables (preventive interventions and risk factors). This cannot constitute a cause and effect relationship due to the retrospective nature of data analyzed; it only supports the association between a number of interventions and risk factors in preventing or predicting PUs. This can guide further research to investigate these interventions and risk factors by employing the same approach used, but in a prospective manner.
- PhD