Search

Background: Treatment burden is the effort required of patients to look after their health and the impact this has on their functioning and wellbeing. It is likely treatment burden changes over time as circumstances change for patients and health services. However, there are a lack of population-level studies of treatment burden change and factors associated with this change over time. Furthermore, there are currently no practical screening tools for treatment burden in time-pressured clinical settings or at population level. Methods and analysis: This is a three-year follow-up of a cross-sectional survey of 723 people with multimorbidity (defined as three or more long-term conditions; LTCs) registered at GP practices in in Dorset, England. The survey will repeat collection of information on treatment burden (using the 10-item Multimorbidity Treatment Burden Questionnaire (MTBQ) and a novel single-item screening tool), sociodemographics, medications, LTCs, health literacy and financial resource, as at baseline. Descriptive statistics will be used to compare change in treatment burden since the baseline survey in 2019 and associations of treatment burden change will be assessed using regression methods. Diagnostic test accuracy metrics will be used to evaluate the single-item treatment burden screening tool using the MTBQ as the gold-standard. Routine primary care data (including demographics, medications, LTCs, and healthcare usage data) will be extracted from medical records for consenting participants. A forward-stepwise, likelihood-ratio logistic regression model building approach will be employed in order to assess the utility of routine data metrics in quantifying treatment burden in comparison to self-reported treatment burden using the MTBQ. Impact: To the authors' knowledge, this will be the first study investigating longitudinal aspects of treatment burden. Findings will improve understanding of the extent to which treatment burden changes over time for people with multimorbidity and factors contributing to this change, as well as allowing better identification of people at risk of high treatment burden. [ABSTRACT FROM AUTHOR]

Abstract Background Rising rates of unplanned admissions among older people are placing unprecedented demand on health services internationally. Unplanned hospital admissions for ill-defined conditions (coded with an R prefix within Chapter XVIII of the International Classification of Diseases-10) have been targeted for admission avoidance strategies, but little is known about these admissions. The aim of this study was to determine the incidence and factors predicting ill-defined (R-coded) hospital admissions of older people and their association with health outcomes. Methods Retrospective analysis of unplanned hospital admissions to general internal and geriatric medicine wards in one hospital over 12 months (2002) with follow-up for 36 months. The study was carried out in an acute teaching hospital in England. The participants were all people aged 65 and over with unplanned hospital admissions to general internal and geriatric medicine. Independent variables included time of admission, residence at admission, route of admission to hospital, age, gender, comorbidity measured by count of diagnoses. Main outcome measures were primary diagnosis (ill-defined versus other diagnostic code), death during the hospital stay, deaths to 36 months, readmissions within 36 months, discharge destination and length of hospital stay. Results Incidence of R-codes at discharge was 21.6%, but was higher in general internal than geriatric medicine (25.6% v 14.1% respectively). Age, gender and co-morbidity were not significant predictors of R-code diagnoses. Admission via the emergency department (ED), out of normal general practitioner (GP) hours, under the care of general medicine and from non-residential care settings increased the risk of receiving R-codes. R-coded patients had a significantly shorter length of stay (5.91 days difference, 95% CI 4.47, 7.35), were less likely to die (hazard ratio 0.71, 95%CI 0.59, 0.85) at any point, but were as likely to be readmitted as other patients (hazard ratio 0.96 (95% CI 0.88, 1.05). Conclusions R-coded diagnoses accounted for 1/5 of emergency admission episodes, higher than anticipated from total English hospital admissions, but comparable with rates reported in similar settings in other countries. Unexpectedly, age did not predict R-coded diagnosis at discharge. Lower mortality and length of stay support the view that these are avoidable admissions, but readmission rates particularly for further R-coded admissions indicate on-going health care needs. Patient characteristics did not predict R-coding, but organisational features, particularly admission via the ED, out of normal GP hours and via general internal medicine, were important and may offer opportunity for admission reduction strategies.