Conectar
To determine the predictors of pressure injuries among residents living in Sri Lankan nursing homes.
A prospective multi-site longitudinal cohort study design.
Semi-structured observations and chart audits were used to gather data on 17 predictors of pressure injury from a consecutive sample of 210 residents (aged ≥ 60 years old) from nine nursing homes in Sri Lanka. Data were collected at baseline and followed up every week until the study endpoint: a new pressure injury or reaching the maximum 12 weeks of data collection, from July to October 2023. Validated semi-structured data collection forms and chart audits were utilised. Binary logistic regression was used to identify the predictors of pressure injuries. Generalised linear mixed models were used to assess the association between predictors and the development of new pressure injuries.
The cumulative incidence of pressure injuries was 17.1% (36/210) during the 12 weeks. The number of medical devices and baseline pressure injuries predicted the development of new pressure injuries. Each additional medical device increased the likelihood of developing a pressure injury by 2.3-fold, and individuals with a baseline pressure injury were 2.1 times more likely to develop a new pressure injury.
Multiple medical devices and baseline pressure injuries are predictors of pressure injury in older residents living in nursing homes.
This study provides evidence of pressure injury predictors among older residents living in nursing homes. Early identification of high-risk residents with an existing pressure injury and those with multiple medical devices is important for nurses and managers at nursing homes. Accurately assessing residents' risk of a pressure injury may result in implementing various preventive strategies that may ultimately help prevent future pressure injuries.
Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) for cohort studies guidelines.
No patient or public contribution.
This paper discusses data errors and offers guidance on data cleaning techniques, with a particular focus on handling missing values and outliers in quantitative datasets.
Methodological discussion.
This paper provides an overview of various techniques for identifying and addressing data anomalies, which can arise from incomplete, noisy, and inconsistent data. These anomalies can significantly affect data quality, leading to biased model parameter estimates and evidence-based decisions. Data cleaning, particularly the appropriate handling of missing values and outliers, is essential to improving data quality before analysis. Data cleaning includes screening for anomalies, diagnosing errors, and applying appropriate corrective measures.
Proper handling of missing values and the identification and correction of outliers are crucial aspects of data cleaning in ensuring data quality and the reliability of statistical analyses. Effective data cleaning enhances the validity and accuracy of research findings for evidence-based decision making that leads to optimal patient outcomes.
The quality of study results depends on how a dataset and its complexities are processed or handled before the analysis. Nursing researchers must use a framework to identify and address important data anomalies and produce reliable results.
This paper describes data cleaning, often overlooked during the data mining process, as a crucial step before conducting data analysis. By addressing missing values and outliers, identifying and fixing data anomalies, and enhancing data quality prior to analysis, data cleaning techniques can produce precise research findings for evidence-based decision making.
In this methodological paper, no new data were generated.
No patient or public contribution.
FreshRSS 