The selfBACK artificial intelligence-based smartphone app can improve low back pain outcome even in patients with high levels of depression or stress

BACKGROUND

selfBACK provides individually tailored self-management support for low back pain (LBP) via an artificial intelligence-based smartphone app. We explore whether those with depressive/stress symptoms can benefit from this technology.

METHODS

Secondary analysis of the selfBACK randomized controlled trial (n = 461). Participants with LBP were randomized to usual care (n = 229), or usual care plus selfBACK (n = 232).

PRIMARY OUTCOME

LBP-related disability (Roland-Morris Disability Questionnaire, RMDQ) over 9 months.

SECONDARY OUTCOMES

global perceived effect (GPE)/pain self-efficacy (PSEQ)/satisfaction/app engagement. Baseline depressive symptoms were measured using the patient health questionnaire (PHQ-8) and stress with the perceived stress scale (PSS). Outcomes stratified by baseline PHQ-8/PSS scores to assess associations across the whole cohort, and intervention versus control groups.

RESULTS

Participants with higher levels of depressive/stress symptoms reported more baseline LBP-related disability (RMDQ 3.1; 1.6 points higher in most vs least depressed/stressed groups respectively); lower self-efficacy (PSEQ 8.1; 4.6 points lower in most vs least depressive/stressed groups respectively). LBP-related disability improved over time; relative risk of improvement in those with greatest depressive/stress symptoms versus nil symptom comparators at 9 months: 0.8 (95% CI: 0.6 to 1.0) and 0.8 (95% CI: 0.7 to 1.0) respectively. No evidence that different baseline levels of depressive/perceived stress symptoms are associated with different RMDQ/GPE/PSEQ outcomes. Whilst participants with higher PHQ-8/PSS were less likely to be satisfied or engage with the app, there was no consistent association among PHQ-8/PSS level, the intervention and outcomes.

CONCLUSIONS

The selfBACK app can improve outcomes even in those with high levels of depressive/stress symptoms and could be recommended for patients with LBP.

SIGNIFICANCE

We have demonstrated that an app supporting the self-management of LBP is helpful, even in those with higher levels of baseline depression and stress symptoms. selfBACK offers an opportunity to support people with LBP and provides clinicians with an additional tool for their patients, even those with depression or high levels of stress. This highlights the potential for digital health interventions for chronic pain.

1310 FRAILTY, LONELINESS AND SOCIAL ISOLATION IN THE UK BIOBANK COHORT

Background

Three challenges for ageing populations are frailty (a state of reduced physiological reserve), social isolation (objective lack of social connections), and loneliness (subjective experience of feeling alone). These are associated with adverse outcomes. This study aims to examine how frailty in combination with loneliness or social isolation is associated with all-cause mortality and hospitalisation rate using data from UK Biobank, a large population-based research cohort.

Methods

502,456 UK Biobank participants were recruited 2006-2010. Baseline data assessed frailty (via two measures: Fried frailty phenotype, Rockwood frailty index), social isolation, and loneliness. Adjusted cox-proportional hazards models assessed association between frailty in combination with loneliness or social isolation and all-cause mortality. Negative binomial regression models assessed hospitalisation rate.

Findings

Frailty, social isolation, and loneliness are common in UK Biobank (frail as per frailty phenotype 3.38%, frail as per frailty index 4.68%, social isolation 9.04%, loneliness 4.75%). Social isolation/loneliness were more common in frailty/pre-frailty. Frailty is associated with increased mortality regardless of social isolation/loneliness. Hazard ratios for frailty (frailty phenotype) were 3.38 (3.11-3.67) with social isolation and 2.89 (2.75-3.05) without social isolation, 2.94 (2.64-3.27) with loneliness and 2.9 (2.76-3.04) without loneliness. Social isolation was associated with increased mortality at all levels of frailty; loneliness only in robust/pre-frail. Frailty was also associated with hospitalisation regardless of social isolation/loneliness. Incidence rate ratios for frailty (frailty phenotype) were 3.93 (3.66-4.23) with social isolation and 3.75 (3.6-3.9) without social isolation, 4.42 (4.04-4.83) with loneliness and 3.69 (3.55-3.83) without loneliness. At all levels frailty, social isolation/loneliness are associated with increased hospitalisation Results were similar using the frailty index definition.

Conclusion

Social isolation is relevant at all levels frailty. Risk of loneliness is more pronounced in those who are robust or pre-frail. Proactive identification of loneliness, regardless of physical health status may provide opportunities for intervention.

1311 FRAILTY IN RANDOMISED CONTROLLED TRIALS FOR DEMENTIA OR MILD COGNITIVE IMPAIRMENT

Background

Frailty and dementia have a bidirectional relationship. However, frailty is rarely reported in clinical trials for dementia and mild cognitive impairment (MCI) which limits assessment of trial applicability. This study aims to use a frailty index (FI) to measure frailty using individual participant data (IPD) from clinical trials for MCI and dementia and to quality the prevalence of frailty and its association with serious adverse events (SAEs) and trial attrition.

Methods

We analysed IPD from three dementia (n=1) and MCI (n=2) trials. An FI comprising physical deficits was created for each trial using baseline IPD. Poisson and logistic regression were used to examine associations with SAEs and attrition, respectively. Estimates were pooled in random effects meta-analysis. Analyses were repeated using an FI incorporating cognitive as well as physical deficits, and results compared.

Results

The mean physical FI was 0.13 and 0.14 in the MCI trials and 0.25 in the dementia trial. Frailty prevalence (FI>0.24) was 5.1%, 5.4% in MCI trials and 55.6% in dementia. After including cognitive deficits, prevalence was similar in MCI (4.6% and 4.9%) but higher in dementia (80.7%). 99th percentile of FI (0.29 in MCI, 0.44 in dementia) was lower than in most general population studies. Frailty was associated with SAEs (physical FI IRR = 1.63 [1.43, 1.87]; physical/cognitive FI IRR = 1.67 [1.45, 1.93]). Frailty was not associated with trial attrition (physical FI OR = 1.18 [0.92, 1.53]; physical/cognitive FI OR = 1.17 [0.92, 1.49]).

Conclusion

Measuring frailty from IPD in dementia and MCI trials is feasible. Severe frailty may be under-represented. Frailty is associated with clinically significant outcomes. Including only physical deficits may underestimate frailty in dementia. Frailty can and should be measured in trials for dementia and MCI, and efforts should be made to facilitate inclusion of people living with frailty.

Systematic review and meta-analysis of prevalence, trajectories, and clinical outcomes for frailty in COPD

This systematic review synthesised measurement and prevalence of frailty in COPD and associations between frailty and adverse health outcomes. We searched Medline, Embase and Web of Science (1 January 2001-8 September 2021) for observational studies in adults with COPD assessing frailty prevalence, trajectories, or association with health-related outcomes. We performed narrative synthesis and random-effects meta-analyses. We found 53 eligible studies using 11 different frailty measures. Most common were frailty phenotype (n = 32), frailty index (n = 5) and Kihon checklist (n = 4). Prevalence estimates varied by frailty definitions, setting, and age (2.6-80.9%). Frailty was associated with mortality (5/7 studies), COPD exacerbation (7/11), hospitalisation (3/4), airflow obstruction (11/14), dyspnoea (15/16), COPD severity (10/12), poorer quality of life (3/4) and disability (1/1). In conclusion, frailty is a common among people with COPD and associated with increased risk of adverse outcomes. Proactive identification of frailty may aid risk stratification and identify candidates for targeted intervention.

Classification of long-term condition patterns in rheumatoid arthritis and associations with adverse health events: a UK Biobank cohort study

PURPOSE

We aimed to classify individuals with RA and ≥2 additional long-term conditions (LTCs) and describe the association between different LTC classes, number of LTCs and adverse health outcomes.

METHODS

We used UK Biobank participants who reported RA (n=5,625) and employed latent class analysis (LCA) to create classes of LTC combinations for those with ≥2 additional LTCs. Cox-proportional hazard and negative binomial regression were used to compare the risk of all-cause mortality, major adverse cardiac events (MACE), and number of emergency hospitalisations over an 11-year follow-up across the different LTC classes and in those with RA plus one additional LTC. Persons with RA without LTCs were the reference group. Analyses were adjusted for demographic characteristics, smoking, BMI, alcohol consumption and physical activity.

RESULTS

A total of 2,566 (46%) participants reported ≥2 LTCs in addition to RA. This involved 1,138 distinct LTC combinations of which 86% were reported by ≤2 individuals. LCA identified 5 morbidity-classes. The distinctive condition in the class with the highest mortality was cancer (class 5; HR 2.66 95%CI (1.91-3.70)). The highest MACE (HR 2.95 95%CI (2.11-4.14)) and emergency hospitalisations (rate ratio 3.01 (2.56-3.54)) were observed in class 3 which comprised asthma, COPD & CHD. There was an increase in mortality, MACE and emergency hospital admissions within each class as the number of LTCs increased.

CONCLUSIONS

The risk of adverse health outcomes in RA varied with different patterns of multimorbidity. The pattern of multimorbidity should be considered in risk assessment and formulating management plans in patients with RA.

Correlations between comorbidities in trials and the community: An individual-level participant data meta-analysis

Background

People with comorbidities are under-represented in randomised controlled trials, and it is unknown whether patterns of comorbidity are similar in trials and the community.

Methods

Individual-level participant data were obtained for 83 clinical trials (54,688 participants) for 16 index conditions from two trial repositories: Yale University Open Data Access (YODA) and the Centre for Global Clinical Research Data (Vivli). Community data (860,177 individuals) were extracted from the Secure Anonymised Information Linkage (SAIL) databank for the same index conditions. Comorbidities were defined using concomitant medications. For each index condition, we estimated correlations between comorbidities separately in trials and community data. For the six commonest comorbidities we estimated all pairwise correlations using Bayesian multivariate probit models, conditioning on age and sex. Correlation estimates from trials with the same index condition were combined into a single estimate. We then compared the trial and community estimates for each index condition.

Results

Despite a higher prevalence of comorbidities in the community than in trials, the correlations between comorbidities were mostly similar in both settings. On comparing correlations between the community and trials, 21% of correlations were stronger in the community, 10% were stronger in the trials and 68% were similar in both. In the community, 5% of correlations were negative, 21% were null, 56% were weakly positive and 18% were strongly positive. Equivalent results for the trials were 11%, 33%, 45% and 10% respectively.

Conclusions

Comorbidity correlations are generally similar in both the trials and community, providing some evidence for the reporting of comorbidity-specific findings from clinical trials.

Multidisciplinary ecosystem to study lifecourse determinants and prevention of early-onset burdensome multimorbidity (MELD-B) - protocol for a research collaboration

Background

Most people living with multiple long-term condition multimorbidity (MLTC-M) are under 65 (defined as 'early onset'). Earlier and greater accrual of long-term conditions (LTCs) may be influenced by the timing and nature of exposure to key risk factors, wider determinants or other LTCs at different life stages. We have established a research collaboration titled 'MELD-B' to understand how wider determinants, sentinel conditions (the first LTC in the lifecourse) and LTC accrual sequence affect risk of early-onset, burdensome MLTC-M, and to inform prevention interventions.

Aim

Our aim is to identify critical periods in the lifecourse for prevention of early-onset, burdensome MLTC-M, identified through the analysis of birth cohorts and electronic health records, including artificial intelligence (AI)-enhanced analyses.

Design

We will develop deeper understanding of 'burdensomeness' and 'complexity' through a qualitative evidence synthesis and a consensus study. Using safe data environments for analyses across large, representative routine healthcare datasets and birth cohorts, we will apply AI methods to identify early-onset, burdensome MLTC-M clusters and sentinel conditions, develop semi-supervised learning to match individuals across datasets, identify determinants of burdensome clusters, and model trajectories of LTC and burden accrual. We will characterise early-life (under 18 years) risk factors for early-onset, burdensome MLTC-M and sentinel conditions. Finally, using AI and causal inference modelling, we will model potential 'preventable moments', defined as time periods in the life course where there is an opportunity for intervention on risk factors and early determinants to prevent the development of MLTC-M. Patient and public involvement is integrated throughout.

Multimorbidity and Blood Pressure Control in Patients Attending Primary Care in Canada

BACKGROUND

There has been conflicting evidence on the association between multimorbidity and blood pressure (BP) control. This study aimed to investigate this associations in people with hypertension attending primary care in Canada, and to assess whether individual long-term conditions are associated with BP control.

METHODS

This was a cross-sectional study in people with hypertension attending primary care in Toronto between January 1, 2017 and December 31, 2019. Uncontrolled BP was defined as systolic BP ≥ 140 mmHg or diastolic BP ≥ 90 mmHg. A list of 11 a priori selected chronic conditions was used to define multimorbidity. Multimorbidity was defined as having ≥1 long-term condition in addition to hypertension. Logistic regression models were used to estimate the association between multimorbidity (or individual long-term conditions) with uncontrolled BP.

RESULTS

A total of 67 385 patients with hypertension were included. They had a mean age of 70, 53.1% were female, 80.6% had multimorbidity, and 35.7% had uncontrolled BP. Patients with multimorbidity had lower odds of uncontrolled BP than those without multimorbidity (adjusted OR = 0.72, 95% CI 0.68-0.76). Among the long-term conditions, diabetes (aOR = 0.73, 95%CI 0.70-0.77), heart failure (aOR = 0.81, 95%CI 0.73-0.91), ischemic heart disease (aOR = 0.74, 95%CI 0.69-0.79), schizophrenia (aOR = 0.79, 95%CI 0.65-0.97), depression/anxiety (aOR = 0.91, 95%CI 0.86-0.95), dementia (aOR = 0.87, 95%CI 0.80-0.95), and osteoarthritis (aOR = 0.89, 95%CI 0.85-0.93) were associated with a lower likelihood of uncontrolled BP.

CONCLUSION

We found that multimorbidity was associated with better BP control. Several conditions were associated with better control, including diabetes, heart failure, ischemic heart disease, schizophrenia, depression/anxiety, dementia, and osteoarthritis.

Using qualitative study designs to understand treatment burden and capacity for self-care among patients with HIV/NCD multimorbidity in South Africa: A methods paper

Background

Low- and middle-income countries (LMICs), including South Africa, are currently experiencing multiple epidemics: HIV and the rising burden of non-communicable diseases (NCDs), leading to different patterns of multimorbidity (the occurrence of two or more chronic conditions) than experienced in high income settings. These adversely affect health outcomes, increase patients' perceived burden of treatment, and impact the workload of self-management. This paper outlines the methods used in a qualitative study exploring burden of treatment among people living with HIV/NCD multimorbidity in South Africa.

Methods

We undertook a comparative qualitative study to examine the interaction between individuals' treatment burden (self-management workload) and their capacity to take on this workload, using the dual lenses of Burden of Treatment Theory (BoTT) and Cumulative Complexity Model (CuCoM) to aid conceptualisation of the data. We interviewed 30 people with multimorbidity and 16 carers in rural Eastern Cape and urban Cape Town between February-April 2021. Data was analysed through framework analysis.

Findings

This paper discusses the methodological procedures considered when conducting qualitative research among people with multimorbidity in low-income settings in South Africa. We highlight the decisions made when developing the research design, recruiting participants, and selecting field-sites. We also explore data analysis processes and reflect on the positionality of the research project and researchers.

Conclusion

This paper illustrates the decision-making processes conducting this qualitative research and may be helpful in informing future research aiming to qualitatively investigate treatment burden among patients in LMICs.

A conceptual framework for characterising lifecourse determinants of multiple long-term condition multimorbidity

Objective

Social, biological and environmental factors in early-life, defined as the period from preconception until age 18, play a role in shaping the risk of multiple long-term condition multimorbidity. However, there is a need to conceptualise these early-life factors, how they relate to each other, and provide conceptual framing for future research on aetiology and modelling prevention scenarios of multimorbidity. We develop a conceptual framework to characterise the population-level domains of early-life determinants of future multimorbidity.

Method

This work was conducted as part of the Multidisciplinary Ecosystem to study Lifecourse Determinants and Prevention of Early-onset Burdensome Multimorbidity (MELD-B) study. The conceptualisation of multimorbidity lifecourse determinant domains was shaped by a review of existing research evidence and policy, and co-produced with public involvement via two workshops.

Results

Early-life risk factors incorporate personal, social, economic, behavioural and environmental factors, and the key domains discussed in research evidence, policy, and with public contributors included adverse childhood experiences, socioeconomics, the social and physical environment, and education. Policy recommendations more often focused on individual-level factors as opposed to the wider determinants of health discussed within the research evidence. Some domains highlighted through our co-production process with public contributors, such as religion and spirituality, health screening and check-ups, and diet, were not adequately considered within the research evidence or policy.

Conclusions

This co-produced conceptualisation can inform research directions using primary and secondary data to investigate the early-life characteristics of population groups at risk of future multimorbidity, as well as policy directions to target public health prevention scenarios of early-onset multimorbidity.

Comorbidity and health-related quality of life in people with a chronic medical condition in randomised clinical trials: An individual participant data meta-analysis

BACKGROUND

Health-related quality of life metrics evaluate treatments in ways that matter to patients, so are often included in randomised clinical trials (hereafter trials). Multimorbidity, where individuals have 2 or more conditions, is negatively associated with quality of life. However, whether multimorbidity predicts change over time or modifies treatment effects for quality of life is unknown. Therefore, clinicians and guideline developers are uncertain about the applicability of trial findings to people with multimorbidity. We examined whether comorbidity count (higher counts indicating greater multimorbidity) (i) is associated with quality of life at baseline; (ii) predicts change in quality of life over time; and/or (iii) modifies treatment effects on quality of life.

METHODS AND FINDINGS

Included trials were registered on the United States trials registry for selected index medical conditions and drug classes, phase 2/3, 3 or 4, had ≥300 participants, a nonrestrictive upper age limit, and were available on 1 of 2 trial repositories on 21 November 2016 and 18 May 2018, respectively. Of 124 meeting these criteria, 56 trials (33,421 participants, 16 index conditions, and 23 drug classes) collected a generic quality of life outcome measure (35 EuroQol-5 dimension (EQ-5D), 31 36-item short form survey (SF-36) with 10 collecting both). Blinding and completeness of follow up were examined for each trial. Using trials where individual participant data (IPD) was available from 2 repositories, a comorbidity count was calculated from medical history and/or prescriptions data. Linear regressions were fitted for the association between comorbidity count and (i) quality of life at baseline; (ii) change in quality of life during trial follow up; and (iii) treatment effects on quality of life. These results were then combined in Bayesian linear models. Posterior samples were summarised via the mean, 2.5th and 97.5th percentiles as credible intervals (95% CI) and via the proportion with values less than 0 as the probability (PBayes) of a negative association. All results are in standardised units (obtained by dividing the EQ-5D/SF-36 estimates by published population standard deviations). Per additional comorbidity, adjusting for age and sex, across all index conditions and treatment comparisons, comorbidity count was associated with lower quality of life at baseline and with a decline in quality of life over time (EQ-5D -0.02 [95% CI -0.03 to -0.01], PBayes > 0.999). Associations were similar, but with wider 95% CIs crossing the null for SF-36-PCS and SF-36-MCS (-0.05 [-0.10 to 0.01], PBayes = 0.956 and -0.05 [-0.10 to 0.01], PBayes = 0.966, respectively). Importantly, there was no evidence of any interaction between comorbidity count and treatment efficacy for either EQ-5D or SF-36 (EQ-5D -0.0035 [95% CI -0.0153 to -0.0065], PBayes = 0.746; SF-36-MCS (-0.0111 [95% CI -0.0647 to 0.0416], PBayes = 0.70 and SF-36-PCS -0.0092 [95% CI -0.0758 to 0.0476], PBayes = 0.631.

CONCLUSIONS

Treatment effects on quality of life did not differ by multimorbidity (measured via a comorbidity count) at baseline-for the medical conditions studied, types and severity of comorbidities and level of quality of life at baseline, suggesting that evidence from clinical trials is likely to be applicable to settings with (at least modestly) higher levels of comorbidity.

TRIAL REGISTRATION

A prespecified protocol was registered on PROSPERO (CRD42018048202).

Associations between a weighted health behaviour score, socioeconomic status, and all-cause mortality in UK Biobank cohort: a prospective analysis

BACKGROUND

Unhealthy behaviours are associated with disproportionate mortality among socioeconomically deprived populations. Previous studies exploring that disproportionate harm do not examine weighted scores, or examine few behaviours. We aimed to create an extended weighted health behaviour score and examine the effect of socioeconomic status on the association between score and all-cause mortality.

METHODS

Data was sourced from the UK Biobank population cohort, recruited in 2006-10. The main exposures included in the analysis were 11 health behaviours (baseline self-report): smoking status, alcohol consumption, physical activity, time spent watching television, sleep duration, added salt in diet, social isolation, intake of red meat, intake of processed meat, intake of oily fish, and intake of fruit and vegetables. Behaviours were classified as healthy or unhealthy according to national guidelines or latest evidence. Socioeconomic deprivation was measured with the Townsend deprivation index. Cox proportional hazard models of health behaviour data were prospectively linked to death registries to examine associations between health behaviours and all-cause mortality. Models were adjusted for demographics and health at baseline. Mortality associated with each behaviour alone was used to determine score weighting. For sensitivity analysis, we explored associations between weighted lifestyle score and all-cause mortality stratified by sex and ethnicity.

FINDINGS

The analysis included 229 107 participants with complete data. Median age was 53 years (IQR 47-60) for 119 634 (52·2%) women and 54 years (47-60) for 109 473 (47·8%) men. Over a median follow-up of 11·9 years (IQR 11.1-12.6), 9379 (4·1%) participants died. Compared with having no unhealthy behaviours, each behaviour was positively associated with all-cause mortality. Smoking (hazard ratio [HR] 2·47 [95% CI 2·25-2·70]) and social isolation (1·69 [1·54-1·86]) were associated with notably higher mortality. A weighted score was created by ascribing one point to each 40% increment in risk (four points for smoking, two points for social isolation, and one points for each of all other behaviours). A dose-response increment for all-cause mortality HR was noted with each additional point of weighted score. Associations were stronger in more deprived tertiles. With least deprived and lowest score as reference, HRs for highest scores were 2·22 (95% CI 1·72-2·86) in the least deprived and 4·10 (3·62-4·65) in the most deprived. An additive interaction between sex and lifestyle score for all-cause mortality was suggested by the data; men had slightly higher HRs at each level of the lifestyle score. However, a statistical test for interaction on a multiplicative scale was not significant. No evidence was found of interaction (either additive or multiplicative) between ethnicity and lifestyle score.

INTERPRETATION

An extended weighted health behaviour score has strong associations with mortality, and associations are stronger in more deprived participants. Weighted health behaviour scores that account for socioeconomic deprivation could convey personalised risk and inform healthy living policy. Further work with adequate numbers of participants from minority ethnic groups is required to make more accurate estimates of mortality associated with a weighted health behaviour score in these populations.

FUNDING

HMEF is supported by a Medical Research Council Clinical Research Training Fellowship (grant number MR/T001585/1), which covered the costs of accessing the data herein.

The presence and impact of multimorbidity clusters on adverse outcomes across the spectrum of kidney function

BACKGROUND

Multimorbidity (the presence of two or more chronic conditions) is common amongst people with chronic kidney disease, but it is unclear which conditions cluster together and if this changes as kidney function declines. We explored which clusters of conditions are associated with different estimated glomerular filtration rates (eGFRs) and studied associations between these clusters and adverse outcomes.

METHODS

Two population-based cohort studies were used: the Stockholm Creatinine Measurements project (SCREAM, Sweden, 2006-2018) and the Secure Anonymised Information Linkage Databank (SAIL, Wales, 2006-2021). We studied participants in SCREAM (404,681 adults) and SAIL (533,362) whose eGFR declined lower than thresholds (90, 75, 60, 45, 30 and 15 mL/min/1.73m²). Clusters based on 27 chronic conditions were identified. We described the most common chronic condition(s) in each cluster and studied their association with adverse outcomes using Cox proportional hazards models (all-cause mortality (ACM) and major adverse cardiovascular events (MACE)).

RESULTS

Chronic conditions became more common and clustered differently across lower eGFR categories. At eGFR 90, 75, and 60 mL/min/1.73m², most participants were in large clusters with no prominent conditions. At eGFR 15 and 30 mL/min/1.73m², clusters involving cardiovascular conditions were larger and were at the highest risk of adverse outcomes. At eGFR 30 mL/min/1.73m², in the heart failure, peripheral vascular disease and diabetes cluster in SCREAM, ACM hazard ratio (HR) is 2.66 (95% confidence interval (CI) 2.31-3.07) and MACE HR is 4.18 (CI 3.65-4.78); in the heart failure and atrial fibrillation cluster in SAIL, ACM HR is 2.23 (CI 2.04 to 2.44) and MACE HR is 3.43 (CI 3.22-3.64). Chronic pain and depression were common and associated with adverse outcomes when combined with physical conditions. At eGFR 30 mL/min/1.73m², in the chronic pain, heart failure and myocardial infarction cluster in SCREAM, ACM HR is 2.00 (CI 1.62-2.46) and MACE HR is 4.09 (CI 3.39-4.93); in the depression, chronic pain and stroke cluster in SAIL, ACM HR is 1.38 (CI 1.18-1.61) and MACE HR is 1.58 (CI 1.42-1.76).

CONCLUSIONS

Patterns of multimorbidity and corresponding risk of adverse outcomes varied with declining eGFR. While diabetes and cardiovascular disease are known high-risk conditions, chronic pain and depression emerged as important conditions and associated with adverse outcomes when combined with physical conditions.

Which combinations of health behaviours are associated with highest risk? An exploration of the UK Biobank population cohort

BACKGROUND

Combinations of unhealthy behaviours are associated with greater mortality than single behaviours, but some combinations might have stronger associations than others. High-risk combinations might be more prevalent among socioeconomically deprived populations. We examined associations between combinations of 11 unhealthy behaviours and mortality and explored socioeconomic distributions of high-risk combinations.

METHODS

We used the UK Biobank population cohort (n=502 459; aged 37-73 years) recruited between 2006 and 2010. Analysis included 229 197 participants with complete data. Main exposures were any combination of smoking, alcohol, physical activity, television time, sleep, added salt, social isolation, intake of red meat, processed meat, oily fish, and fruit and vegetables (each classified as healthy or unhealthy via guidelines or latest evidence). Townsend index was used to explore socioeconomic distribution. Cox proportional hazard models were used to examine associations between behaviours and all-cause mortality. Models were adjusted for demographic, health, and socioeconomic factors.

FINDINGS

Over a median follow-up of 11·6 years, 9739 (4·2%) of 229 197 participants died. From 716 unique combinations, 77 (11%) were associated with mortality with hazard ratios (HRs) greater than that for smoking alone (2·31 [95% CI 2·11-2·53]); HRs ranged from 9·44 to 2·34. Of these 77 high-risk combinations, smoking featured in 61 (79%), low fruit and vegetables in 45 (58%), and low oily fish in 41 (53%). All combinations featuring social isolation (18 [23%] of 77) had HRs greater than 3·00. Participants with high-risk combinations had greater deprivation scores than those with no unhealthy behaviours. Median deprivation scores of the ten highest risk combinations ranged from -2·0 to 2·1, whereas for participants with no unhealthy behaviours the score was -2·5. Examining women and men separately resulted in similar findings. Examination of ethnic differences was severely limited by small numbers of participants in minority ethnic groups.

INTERPRETATION

Many unique unhealthy behaviour combinations are strongly associated with mortality and high-risk combinations are more prevalent among more deprived populations than among more affluent populations. Exploring unique combinations of a wide range of health behaviours can identify high-risk populations. Future work with adequately sampled minority ethnic groups is required to examine high-risk combinations by ethnicity. Supporting healthy living in deprived populations, including tackling structural barriers to health, could address a wide range of health behaviours in combination.

FUNDING

UK Medical Research Council.

Assessing trial representativeness using serious adverse events: an observational analysis using aggregate and individual-level data from clinical trials and routine healthcare data

BACKGROUND

The applicability of randomised controlled trials of pharmacological agents to older people with frailty/multimorbidity is often uncertain, due to concerns that trials are not representative. However, assessing trial representativeness is challenging and complex. We explore an approach assessing trial representativeness by comparing rates of trial serious adverse events (SAE) to rates of hospitalisation/death in routine care.

METHODS

This was an observational analysis of individual (125 trials, n=122,069) and aggregate-level drug trial data (483 trials, n=636,267) for 21 index conditions compared to population-based routine healthcare data (routine care). Trials were identified from ClinicalTrials.gov . Routine care comparison from linked primary care and hospital data from Wales, UK (n=2.3M). Our outcome of interest was SAEs (routinely reported in trials). In routine care, SAEs were based on hospitalisations and deaths (which are SAEs by definition). We compared trial SAEs in trials to expected SAEs based on age/sex standardised routine care populations with the same index condition. Using IPD, we assessed the relationship between multimorbidity count and SAEs in both trials and routine care and assessed the impact on the observed/expected SAE ratio additionally accounting for multimorbidity.

RESULTS

For 12/21 index conditions, the pooled observed/expected SAE ratio was <1, indicating fewer SAEs in trial participants than in routine care. A further 6/21 had point estimates <1 but the 95% CI included the null. The median pooled estimate of observed/expected SAE ratio was 0.60 (95% CI 0.55-0.64; COPD) and the interquartile range was 0.44 (0.34-0.55; Parkinson's disease) to 0.87 (0.58-1.29; inflammatory bowel disease). Higher multimorbidity count was associated with SAEs across all index conditions in both routine care and trials. For most trials, the observed/expected SAE ratio moved closer to 1 after additionally accounting for multimorbidity count, but it nonetheless remained below 1 for most.

CONCLUSIONS

Trial participants experience fewer SAEs than expected based on age/sex/condition hospitalisation and death rates in routine care, confirming the predicted lack of representativeness. This difference is only partially explained by differences in multimorbidity. Assessing observed/expected SAE may help assess the applicability of trial findings to older populations in whom multimorbidity and frailty are common.

One size does not fit all: Participants' experiences of the selfBACK app to support self-management of low back pain-a qualitative interview study

BACKGROUND

Low back pain (LBP) is one of the most common reasons for disability globally. Digital interventions are a promising means of supporting people to self-manage LBP, but implementation of digital interventions has been suboptimal. An artificial intelligence-driven app, selfBACK, was developed to support self-management of LBP as an adjunct to usual care. To better understand the process of implementation from a participant perspective, we qualitatively explored factors influencing embedding, integrating, and sustaining engagement with the selfBACK app, and the self-perceived effects, acceptability, and satisfaction with the selfBACK app.

METHODS

Using a qualitative interview study and an analytic framework approach underpinned by Normalization Process Theory (NPT), we investigated the experiences of patients who participated in the selfBACK randomized controlled trial (RCT). Interviews focused on the motivation to participate in the RCT, experiences of using the selfBACK app, and views about future intended use and potential of using digital health interventions for self-management of LBP. Participants were purposively sampled to represent diversity in age, sex, and implementation reflected by a proxy measure of number of app-generated self-management plans during the first three months of RCT participation.

RESULTS

Twenty-six participants aged 21-78, eleven females and fifteen men, with two to fourteen self-management plans, were interviewed between August 2019 and April 2020. A broad range of factors influencing implementation of selfBACK within all constructs of NPT were identified. Key facilitating factors were preferences and beliefs favoring self-management, a friendly, motivational, and reassuring supporter, tailoring and personalization, convenience and ease of use, trustworthiness, perceiving benefits, and tracking achievements. Key impeding factors were preferences and beliefs not favoring self-management, functionality issues, suboptimal tailoring and personalization, insufficient time or conflicting life circumstances, not perceiving benefits, and insufficient involvement of health care practitioners. Self-perceived effects on pain and health, behavior/attitude, and gaining useful knowledge varied by participant.

CONCLUSIONS

The high prevalence of LBP globally coupled with the advantages of providing help through an app offers opportunities to help countless people. A range of factors should be considered to facilitate implementation of self-management of LBP or similar pain conditions using digital health tools.

Frailty in people with rheumatoid arthritis: a systematic review of observational studies

Background: Frailty, an age-related decline in physiological reserve, is an increasingly important concept in the management of chronic diseases. The implications of frailty in people with rheumatoid arthritis are not well understood. We undertook a systematic review to assess prevalence of frailty in people with rheumatoid arthritis, and the relationship between frailty and disease activity or clinical outcomes. Methods: We searched four electronic databases (January 2001 to April 2021) for observational studies assessing the prevalence of frailty (any frailty measure) in adults (≥18 years) with rheumatoid arthritis, or analysing the relationship between frailty and disease activity or clinical outcomes (e.g. quality of life, hospitalisation or mortality) in people with rheumatoid arthritis. Study quality was assessed using an adapted Newcastle-Ottawa Scale. Screening, quality assessment and data extraction were performed independently by two reviewers. We used narrative synthesis. Results: We identified 17 analyses, from 14 different populations. 15/17 were cross-sectional. Studies used 11 different measures of frailty. Frailty prevalence ranged from 10% (frailty phenotype) to 36% (comprehensive rheumatologic assessment of frailty) in general adult populations with rheumatoid arthritis. In younger populations (<60 or <65 years) prevalence ranged from 2.4% (frailty phenotype) to 19.9% (Kihon checklist) while in older populations (>60 or >65) prevalence ranged from 31.2% (Kihon checklist) to 55% (Geriatric 8 tool). Frailty was cross-sectionally associated with higher disease activity (10/10 studies), lower physical function (7/7 studies) and longer disease duration (2/5 studies), and with hospitalization and osteoporotic fractures (1/1 study, 3.7 years follow-up). Conclusion: Frailty is common in rheumatoid arthritis, including those aged <65 years, and is associated with a range of adverse features. However, these is heterogeneity in how frailty is measured. We found few longitudinal studies making the impact of frailty on clinical outcomes over time and the extent to which frailty is caused by rheumatoid arthritis unclear.

Effect of age, sex, and morbidity count on trial attrition: meta-analysis of individual participant level data from phase 3/4 industry funded clinical trials

Objectives

To estimate the association between individual participant characteristics and attrition from randomised controlled trials.

Design

Meta-analysis of individual participant level data (IPD).

Data sources

Clinical trial repositories (Clinical Study Data Request and Yale University Open Data Access).

Eligibility criteria for selecting studies

Eligible phase 3 or 4 trials identified according to prespecified criteria (PROSPERO CRD42018048202).

Main outcome measures

Association between comorbidity count (identified using medical history or concomitant drug treatment data) and trial attrition (failure for any reason to complete the final trial visit), estimated in logistic regression models and adjusted for age and sex. Estimates were meta-analysed in bayesian linear models, with partial pooling across index conditions and drug classes.

Results

In 92 trials across 20 index conditions and 17 drug classes, the mean comorbidity count ranged from 0.3 to 2.7. Neither age nor sex was clearly associated with attrition (odds ratio 1.04, 95% credible interval 0.98 to 1.11; and 0.99, 0.93 to 1.05, respectively). However, comorbidity count was associated with trial attrition (odds ratio per additional comorbidity 1.11, 95% credible interval 1.07 to 1.14). No evidence of non-linearity (assessed via a second order polynomial) was seen in the association between comorbidity count and trial attrition, with minimal variation across drug classes and index conditions. At a trial level, an increase in participant comorbidity count has a minor impact on attrition: for a notional trial with high level of attrition in individuals without comorbidity, doubling the mean comorbidity count from 1 to 2 translates to an increase in trial attrition from 29% to 31%.

Conclusions

Increased comorbidity count, irrespective of age and sex, is associated with a modest increased odds of participant attrition. The benefit of increased generalisability of including participants with multimorbidity seems likely to outweigh the disadvantages of increased attrition.

Multimorbidity

Multimorbidity (two or more coexisting conditions in an individual) is a growing global challenge with substantial effects on individuals, carers and society. Multimorbidity occurs a decade earlier in socioeconomically deprived communities and is associated with premature death, poorer function and quality of life and increased health-care utilization. Mechanisms underlying the development of multimorbidity are complex, interrelated and multilevel, but are related to ageing and underlying biological mechanisms and broader determinants of health such as socioeconomic deprivation. Little is known about prevention of multimorbidity, but focusing on psychosocial and behavioural factors, particularly population level interventions and structural changes, is likely to be beneficial. Most clinical practice guidelines and health-care training and delivery focus on single diseases, leading to care that is sometimes inadequate and potentially harmful. Multimorbidity requires person-centred care, prioritizing what matters most to the individual and the individual's carers, ensuring care that is effectively coordinated and minimally disruptive, and aligns with the patient's values. Interventions are likely to be complex and multifaceted. Although an increasing number of studies have examined multimorbidity interventions, there is still limited evidence to support any approach. Greater investment in multimorbidity research and training along with reconfiguration of health care supporting the management of multimorbidity is urgently needed.

Frailty in COPD: an analysis of prevalence and clinical impact using UK Biobank

BACKGROUND

Frailty, a state of reduced physiological reserve, is common in people with chronic obstructive pulmonary disease (COPD). Frailty can occur at any age; however, the implications in younger people (eg, aged <65 years) with COPD are unclear. We assessed the prevalence of frailty in UK Biobank participants with COPD; explored relationships between frailty and forced expiratory volume in 1 second (FEV1) and quantified the association between frailty and adverse outcomes.

METHODS

UK Biobank participants (n=3132, recruited 2006-2010) with COPD aged 40-70 years were analysed comparing two frailty measures (frailty phenotype and frailty index) at baseline. Relationship with FEV1 was assessed for each measure. Outcomes were mortality, major adverse cardiovascular event (MACE), all-cause hospitalisation, hospitalisation with COPD exacerbation and community COPD exacerbation over 8 years of follow-up.

RESULTS

Frailty was common by both definitions (17% frail using frailty phenotype, 28% moderate and 4% severely frail using frailty index). The frailty phenotype, but not the frailty index, was associated with lower FEV1. Frailty phenotype (frail vs robust) was associated with mortality (HR 2.33; 95% CI 1.84 to 2.96), MACE (2.73; 1.66 to 4.49), hospitalisation (incidence rate ratio 3.39; 2.77 to 4.14) hospitalised exacerbation (5.19; 3.80 to 7.09) and community exacerbation (2.15; 1.81 to 2.54), as was frailty index (severe vs robust) (mortality (2.65; 95% CI 1.75 to 4.02), MACE (6.76; 2.68 to 17.04), hospitalisation (3.69; 2.52 to 5.42), hospitalised exacerbation (4.26; 2.37 to 7.68) and community exacerbation (2.39; 1.74 to 3.28)). These relationships were similar before and after adjustment for FEV1.

CONCLUSION

Frailty, regardless of age or measure, identifies people with COPD at risk of adverse clinical outcomes. Frailty assessment may aid risk stratification and guide-targeted intervention in COPD and should not be limited to people aged >65 years.

963 THE IMPACT OF AGE, SEX AND MORBIDITY COUNT ON EARLY TERMINATION: A META-ANALYSIS OF INDIVIDUAL PATIENT DATA FROM CLINICAL TRIALS

Introduction

Multimorbidity is found in around half of people with any long-term condition but is substantially less common in randomised controlled trials (‘trials’). Multimorbidity may diminish a participant’s ability to complete a trial. However, empirical estimates of the association between individual patient characteristics and early termination are lacking.

Method

Individual patient-level data were obtained from Phase 3/4 trials contained within two clinical trial repositories. Eligible trials for inclusion were identified according to pre-specified criteria (PROSPERO CRD42018048202). Within each trial, the association between morbidity count and early termination (failure for any reason to complete the final trial visit) was estimated in logistic regression models, adjusting for age and sex. These estimates were meta-analysed in Bayesian linear models, with partial pooling across index conditions and drug classes. Using these estimates, the impact of morbidity count on early termination was modelled for a set of notional trials.

Results

In 92 trials across 20 index conditions and 17 drug classes, the mean morbidity count ranged from 0.3–2.7. Neither age nor sex was associated with early termination (OR 1.04, 95% CI 0.98–1.11; OR 1.00, 95% PI 0.95–1.07 respectively). Morbidity count was associated with early termination (OR per additional morbidity: 1.11, 95% CI: 1.07 to 1.14). There was no evidence of non-linearity in the association between morbidity count and early termination, and there was minimal variation across drug classes and index conditions. For a notional trial with high level of early termination in individuals without multimorbidity, doubling the mean morbidity count from 1 to 2 increases risk of early termination from 29% to 31%.

Conclusion

Multimorbidity, irrespective of age and sex, is associated with a relatively modest increased odds of early termination of trial participation. The benefit of increased generalisability of trials by including patients with multimorbidity appears likely to outweigh the disadvantages of lower retention.

Translational framework for implementation evaluation and research: Protocol for a qualitative systematic review of studies informed by Normalization Process Theory (NPT) [version 1; peer review: 2 approved]

BACKGROUND

Normalization Process Theory (NPT) identifies mechanisms that have been demonstrated to play an important role in implementation processes. It is now widely used to inform feasibility, process evaluation, and implementation studies in healthcare and other areas of work. This qualitative synthesis of NPT studies aims to better understand how NPT explains observed and reported implementation processes, and to explore the ways in which its constructs explain the implementability, enacting and sustainment of complex healthcare interventions.

METHODS

We will systematically search Scopus, PubMed and Web of Science databases and use the Google Scholar search engine for citations of key papers in which NPT was developed. This will identify English language peer-reviewed articles in scientific journals reporting (a) primary qualitative or mixed methods studies; or, (b) qualitative or mixed methods evidence syntheses in which NPT was the primary analytic framework. Studies may be conducted in any healthcare setting, published between June 2006 and 31 December 2021. We will perform a qualitative synthesis of included studies using two parallel methods: (i) directed content analysis based on an already developed coding manual; and (ii) unsupervised textual analysis using Leximancer® topic modelling software.

OTHER

We will disseminate results of the review using peer reviewed publications, conference and seminar presentations, and social media (Facebook and Twitter) channels. The primary source of funding is the National Institute for Health Research ARC North Thames. No human subjects or personal data are involved and no ethical issues are anticipated.

The association between a lifestyle score, socioeconomic status, and COVID-19 outcomes within the UK Biobank cohort

BACKGROUND

Infection with SARS-CoV-2 virus (COVID-19) impacts disadvantaged groups most. Lifestyle factors are also associated with adverse COVID-19 outcomes. To inform COVID-19 policy and interventions, we explored effect modification of socioeconomic-status (SES) on associations between lifestyle and COVID-19 outcomes.

METHODS

Using data from UK-Biobank, a large prospective cohort of 502,536 participants aged 37-73 years recruited between 2006 and 2010, we assigned participants a lifestyle score comprising nine factors. Poisson regression models with penalised splines were used to analyse associations between lifestyle score, deprivation (Townsend), and COVID-19 mortality and severe COVID-19. Associations between each exposure and outcome were examined independently before participants were dichotomised by deprivation to examine exposures jointly. Models were adjusted for sociodemographic/health factors.

RESULTS

Of 343,850 participants (mean age > 60 years) with complete data, 707 (0.21%) died from COVID-19 and 2506 (0.76%) had severe COVID-19. There was evidence of a nonlinear association between lifestyle score and COVID-19 mortality but limited evidence for nonlinearity between lifestyle score and severe COVID-19 and between deprivation and COVID-19 outcomes. Compared with low deprivation, participants in the high deprivation group had higher risk of COVID-19 outcomes across the lifestyle score. There was evidence for an additive interaction between lifestyle score and deprivation. Compared with participants with the healthiest lifestyle score in the low deprivation group, COVID-19 mortality risk ratios (95% CIs) for those with less healthy scores in low versus high deprivation groups were 5.09 (1.39-25.20) and 9.60 (4.70-21.44), respectively. Equivalent figures for severe COVID-19 were 5.17 (2.46-12.01) and 6.02 (4.72-7.71). Alternative SES measures produced similar results.

CONCLUSIONS

Unhealthy lifestyles are associated with higher risk of adverse COVID-19, but risks are highest in the most disadvantaged, suggesting an additive influence between SES and lifestyle. COVID-19 policy and interventions should consider both lifestyle and SES. The greatest public health benefit from lifestyle focussed COVID-19 policy and interventions is likely to be seen when greatest support for healthy living is provided to the most disadvantaged groups.

Experiences of multimorbidity in urban and rural Malawi: An interview study of burdens of treatment and lack of treatment

Multimorbidity (presence of ≥2 long term conditions (LTCs)) is a growing global health challenge, yet we know little about the experiences of those living with multimorbidity in low- and middle-income countries (LMICs). We therefore explore: 1) experiences of men and women living with multimorbidity in urban and rural Malawi including their experiences of burden of treatment and 2) examine the utility of Normalization Process Theory (NPT) and Burden of Treatment Theory (BOTT) for structuring analytical accounts of these experiences. We conducted in depth, semi-structured interviews with 32 people in rural (n = 16) and urban settings (n = 16); 16 males, 16 females; 15 under 50 years; and 17 over 50 years. Data were analysed thematically and then conceptualised through the lens of NPT and BOTT. Key elements of burden of treatment identified included: coming to terms with and gaining an understanding of life with multimorbidity; dealing with resulting disruptions to family life; the work of seeking family and community support; navigating healthcare systems; coping with lack of continuity of care; enacting self-management advice; negotiating medical advice; appraising treatments; and importantly, dealing with the burden of lack of treatments/services. Poverty and inadequate healthcare provision constrained capacity to deal with treatment burden while supportive social and community networks were important enabling features. Greater access to health information/education would lessen treatment burden as would better resourced healthcare systems and improved standards of living. Our work demonstrates the utility of NPT and BOTT for aiding conceptualisation of treatment burden issues in LMICs but our findings highlight that 'lack' of access to treatments or services is an important additional burden which must be integrated in accounts of treatment burden in LMICs.

Frailty in rheumatoidrmdopen-2021-002111 arthritis and its relationship with disease activity, hospitalisation and mortality: a longitudinal analysis of the Scottish Early Rheumatoid Arthritis cohort and UK Biobank

OBJECTIVE

To assess the prevalence of frailty in rheumatoid arthritis (RA) and its association with baseline and longitudinal disease activity, all-cause mortality and hospitalisation.

PARTICIPANTS

People with RA identified from the Scottish Early Rheumatoid Arthritis (SERA) inception cohort (newly diagnosed, mean age 58.2 years) and UK Biobank (established disease identified using diagnostic codes, mean age 59 years). Frailty was quantified using the frailty index (both datasets) and frailty phenotype (UK Biobank only). Disease activity was assessed using Disease Activity Score in 28 joints (DAS28) in SERA. Associations between baseline frailty and all-cause mortality and hospitalisation was estimated after adjusting for age, sex, socioeconomic status, smoking and alcohol, plus DAS28 in SERA.

RESULTS

Based on the frailty index, frailty was common in SERA (12% moderate, 0.2% severe) and UK Biobank (20% moderate, 3% severe). In UK Biobank, 23% were frail using frailty phenotype. Frailty index was associated with DAS28 in SERA, as well as age and female sex in both cohorts. In SERA, as DAS28 lessened over time with treatment, mean frailty index also decreased. The frailty index was associated with all-cause mortality (HR moderate/severe frailty vs robust 4.14 (95% CI 1.49 to 11.51) SERA, 1.68 (95% CI 1.26 to 2.13) UK Biobank) and unscheduled hospitalisation (incidence rate ratio 2.27 (95% CI 1.45 to 3.57) SERA 2.74 (95% CI 2.29 to 3.29) UK Biobank). In UK Biobank, frailty phenotype also associated with mortality and hospitalisation.

CONCLUSION

Frailty is common in early and established RA and associated with hospitalisation and mortality. Frailty in RA is dynamic and, for some, may be ameliorated through controlling disease activity in early disease.