Dataset Information

The use of linked routine data to optimise calculation of the Hospital Frailty Risk Score on the basis of previous hospital admissions: a retrospective observational cohort study

ABSTRACT: Summary

Background

The Hospital Frailty Risk Score (HFRS) has been widely but inconsistently applied in published studies, particularly in how diagnostic information recorded in previous hospital admissions is used in its construction. We aimed to assess how many previous admissions should be considered when constructing the HFRS and the influence of frailty risk on long length of stay, in-hospital mortality, and 30-day readmission.

Methods

This is a retrospective observational cohort study of patients aged 75 years or older who had at least one emergency admission to any of 49 hospital sites in the Yorkshire and Humber region of England, UK. We constructed multiple versions of the HFRS for each patient, each form incorporating diagnostic data from progressively more previous admissions in its construction within a 1-year or 2-year window. We assessed the ability of each form of the HFRS to predict long length of stay (>10 days), in-hospital death, and 30-day readmission.

Findings

Between April 1, 2013, and March 31, 2017, 282?091 patients had 675?155 hospital admissions. Regression analyses assessing the different constructions of HFRS showed that the form constructed with diagnostic information recorded in the current and previous two admissions within the preceding 2 years performed best for predicting all three outcomes. Under this construction, 263?432 (39·0%) of 674?615 patient admissions were classified as having low frailty risk, for whom 33?333 (12·7%) had a long length of stay, 10?145 (3·9%) died in hospital, and 45?226 (17·2%) were readmitted within 30 days. By contrast with those patients with low frailty risk, for those with intermediate frailty risk, the probability was 2·5-times higher (95% CI 2·4 to 2·6) for long length of stay, 2·17-times higher (2·1 to 2·2) for in-hospital death, and 0·7% higher (0·5 to 1) for readmission. For patients with high frailty risk, the probability was 4·3-times higher (4·2 to 4·5) for long length of stay, 2·48-times higher (2·4 to 2·6) for in-hospital death, and ?1% (?1·2 to ?0·5) lower for readmission than those with low frailty risk. The intermediate and high frailty risk categories were more important predictors of long length of stay than any of the other rich set of control variables included in our analysis. These categories also proved to be important predictors of in-hospital mortality, with only the Charlson Comorbidity Index offering greater predictive power.

Interpretation

We recommend constructing the HFRS with diagnostic information from the current admission and from the previous two admissions in the preceding 2 years. This HFRS form was a powerful predictor of long length of stay and in-hospital mortality, but less so of emergency readmissions.

Funding

National Institute of Health Research.

SUBMITTER: Street A

PROVIDER: S-EPMC7934406 | biostudies-literature | 2021 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundMost claims-based frailty instruments have been designed for group stratification of older populations according to the risk of adverse health outcomes and not frailty itself. We aimed to develop and validate a tool based on one-year hospital discharge data for stratification on Fried's frailty phenotype (FP).MethodsWe used a three-stage development/validation approach. First, we created a clinical knowledge-driven electronic frailty score (eFS) calculated as the number of deficient organs/systems among 18 critical ones identified from the International Statistical Classification of Diseases and Related Problems, 10th Revision (ICD-10) diagnoses coded in the year before FP assessment. Second, for eFS development and internal validation, we linked individual records from the Lc65+ cohort database to inpatient discharge data from Lausanne University Hospital (CHUV) for the period 2004-2015. The development/internal validation sample included community-dwelling, non-institutionalised residents of Lausanne (Switzerland) recruited in the Lc65+ cohort in three waves (2004, 2009, and 2014), aged 65-70 years at enrolment, and hospitalised at the CHUV at least once in the year preceding the FP assessment. Using this sample, we selected the best performing model for predicting the dichotomised FP, with the eFS or ICD-10-based variables as predictors. Third, we conducted an external validation using 2016 Swiss nationwide hospital discharge data and compared the performance of the eFS model in predicting 13 adverse outcomes to three models relying on well-designed and validated claims-based scores (Claims-based Frailty Index, Hospital Frailty Risk Score, Dr Foster Global Frailty Score).FindingsIn the development/internal validation sample (n = 469), 14·3% of participants (n = 67) were frail. Among 34 models tested, the best-subsets logistic regression model with four predictors (age and sex at FP assessment, time since last hospital discharge, eFS) performed best in predicting the dichotomised FP (area under the curve=0·71; F1 score=0·39) and one-year adverse health outcomes. On the external validation sample (n = 54,815; 153 acute care hospitals), the eFS model demonstrated a similar performance to the three other claims-based scoring models. According to the eFS model, the external validation sample showed an estimated prevalence of 56·8% (n = 31,135) of frail older inpatients at admission.InterpretationThe eFS model is an inexpensive, transportable and valid tool allowing reliable group stratification and individual prioritisation for comprehensive frailty assessment and may be applied to both hospitalised and community-dwelling older adults.FundingThe study received no external funding.

Project description:IntroductionNeurological conditions are a major and increasing cause of hospitalisation among children and young people, but little is known about the impact of neurological conditions on hospital services in England, nor the factors that influence length of stay and bed days per year.ObjectivesTo quantify the hospital usage in children and young people related to neurological conditions, trends over time and variation by ethnicity and deprivation status.MethodsAn ICD10 coding framework identified a cohort of individuals aged 0-19 years with neurological conditions from linked routinely collected healthcare data from England (The Hospital Episode Statistics Admitted Patient Care dataset), from 1 April 2003 to 31 March 2015. Linked outpatient and accident and emergency data were used to supplement missing demographic data. Length of stay and bed days per year per person were calculated. These were separately modelled using random intercept multivariable negative binomial regressions with gender, age, ethnic group, diagnostic group, region of residence and deprivation category as predictors.Results524,442 individuals were identified over the study period, increasing from 49,928 in 2003/04 to 102,840 in 2014/15. Neurological conditions account for 8.8% of inpatient bed days in the 0-14 year old age group. Length of stay and bed days per year vary primarily by age group - e.g. Under 1 year olds had 1.85 times (95%CI 1.83-1.86%) longer stays and over double (2.36 times, 95%CI 2.34-2.37 times) the number of bed days per person per year compared to 5 to 9 year olds - and main diagnostic group, with smaller variations by ethnic group, deprivation and region.ConclusionsNeurological conditions in children and young people have a significant and increasing impact on the NHS in England. Falls in length of stay and bed days per person are more than offset by increasing numbers of children and young people with neurological diagnoses. Variations in length of stay and bed days per year by diagnostic group, ethnic group, age group, deprivation category and region should be taken into account in resource planning.

Project description:BACKGROUND:Administrative data are used to examine variation in thirty-day mortality across health services in several jurisdictions. Hospital performance measurement may be error-prone as information about disease severity is not typically available in routinely collected data to incorporate into case-mix adjusted analyses. Using ischaemic stroke as a case study, we tested the extent to which accounting for disease severity impacts on hospital performance assessment. METHODS:We linked all recorded ischaemic stroke admissions between July, 2011 and June, 2014 to death registrations and a measure of stroke severity obtained at first point of patient contact with health services, across New South Wales, Australia's largest health service jurisdiction. Thirty-day hospital standardised mortality ratios were adjusted for either comorbidities, as is typically done, or for both comorbidities and stroke severity. The impact of stroke severity adjustment on mortality ratios was determined using 95% and 99% control limits applied to funnel plots and by calculating the change in rank order of hospital risk adjusted mortality rates. RESULTS:The performance of the stroke severity adjusted model was superior to incorporating comorbidity burden alone (c-statistic = 0.82 versus 0.75; N = 17,700 patients, 176 hospitals). Concordance in outlier classification was 89% and 97% when applying 95% or 99% control limits to funnel plots, respectively. The sensitivity rates of outlier detection using comorbidity adjustment compared with gold-standard severity and comorbidity adjustment was 74% and 83% with 95% and 99% control limits, respectively. Corresponding positive predictive values were 74% and 91%. Hospital rank order of risk adjusted mortality rates shifted between 0 to 22 places with severity adjustment (Median = 4.0, Inter-quartile Range = 2-7). CONCLUSIONS:Rankings of mortality rates varied widely depending on whether stroke severity was taken into account. Funnel plots yielded largely concordant results irrespective of severity adjustment and may be sufficiently accurate as a screening tool for assessing hospital performance.

Project description:BackgroundDespite recent advances, mortality in children with Down syndrome remains five times higher than in the general population. This study aims to describe the burden, patterns and causes of hospital admissions in infants with Down syndrome, and compare this with infants without Down syndrome in a population-based cohort.MethodsThis study used data from the Wales Electronic Cohort for Children, a cohort of all children born in Wales between 1990 and 2012. The cohort was generated from routine administrative data, linked to create an anonymised data set within the Secure Anonymised Information Linkage databank. This analysis is based on all infants born between January 2003 and January 2012 who were followed to their first birthday, a move out of Wales, death, or until 31 October 2012 (end of follow-up). Infants with Down syndrome were identified using the Congenital Anomaly Register and Information Service in Wales. Multivariable Cox regression was used to compare the time to first hospital admission. Admission codes were used to identify the commonest indications for hospitalisation and to determine the presence of other congenital anomalies.ResultsWe included 324 060 children, 356 of whom had Down syndrome. Of infants with Down syndrome, 80.3% had at least one hospital inpatient admission during the first year of life, compared with 32.9% of infants without Down syndrome. These first admissions were earlier [median of 6 days interquartile range (IQR) (3, 72) compared with 45 days [IQR 6, 166)] and longer [median of 4 days (IQR 1, 15) compared with 1 day (IQR 0, 3)] than in infants without Down syndrome. The most common causes of admissions were congenital abnormalities, respiratory diseases, conditions originating in the perinatal period and infectious diseases. The presence of other congenital abnormalities increased hospitalisations in all infants, but more so in infants with Down syndrome who spent a median of 21 days in hospital (IQR 11, 47) during their first year of life.ConclusionInfants with Down syndrome are at high risk for early, more frequent and longer hospital admissions. Congenital heart disease and respiratory infections remain a major burden in this population. More research is needed to understand how to better manage these conditions particularly in the first month of life when most admissions occur.