ABSTRACT
Collecting data for an individual participant data meta-analysis (IPDMA) project can be time consuming and resource intensive and could still have insufficient power to answer the question of interest. Therefore, researchers should consider the power of their planned IPDMA before collecting IPD. Here we propose a method to estimate the power of a planned IPDMA project aiming to synthesise multiple cohort studies to investigate the (unadjusted or adjusted) effects of potential prognostic factors for a binary outcome. We consider both binary and continuous factors and provide a three-step approach to estimating the power in advance of collecting IPD, under an assumption of the true prognostic effect of each factor of interest. The first step uses routinely available (published) aggregate data for each study to approximate Fisher's information matrix and thereby estimate the anticipated variance of the unadjusted prognostic factor effect in each study. These variances are then used in step 2 to estimate the anticipated variance of the summary prognostic effect from the IPDMA. Finally, step 3 uses this variance to estimate the corresponding IPDMA power, based on a two-sided Wald test and the assumed true effect. Extensions are provided to adjust the power calculation for the presence of additional covariates correlated with the prognostic factor of interest (by using a variance inflation factor) and to allow for between-study heterogeneity in prognostic effects. An example is provided for illustration, and Stata code is supplied to enable researchers to implement the method.
ABSTRACT
OBJECTIVES: Colchicine is commonly used to prevent flares when starting urate-lowering therapy for gout. Patients with gout are frequently concurrently prescribed other medications (such as statins) that may interact with colchicine, increasing the risk of adverse events. The aim of this study was to describe potential prognostic factors for adverse events in patients prescribed colchicine when initiating allopurinol. METHODS: We conducted a retrospective cohort study in linked UK Clinical Practice Research Datalink and Hospital Episode Statistics datasets. Adults initiating allopurinol for gout with colchicine (01/04/1997-30/11/2016) were included. Potential prognostic factors were defined, and the likelihood of adverse events, including diarrhoea, nausea or vomiting, myocardial infarction (MI), neuropathy, myalgia, myopathy, rhabdomyolysis, and bone marrow suppression, were estimated. RESULTS: From 01/04/1997-30/11/2016, 13 945 people with gout initiated allopurinol with colchicine prophylaxis (mean age 63.9 (SD 14.7) years, 78.2% male). One quarter (26%, 95% CI 25% to 27%) were prescribed ≥1 potentially interacting medicines, most commonly statins (21%, 95% CI 20% to 22%). Statins were not associated with increased adverse events, although other drugs were associated with some adverse outcomes. Diarrhoea and MI were associated with more comorbidities and more severe CKD. CONCLUSION: People were given colchicine prophylaxis despite commonly having preexisting prescriptions for medications with potential to interact with colchicine. Adverse events were more common in people who had more comorbidities and certain potentially interacting medications. Our findings will provide much-needed information about prognostic factors for colchicine-related adverse events that can inform treatment decisions about prophylaxis when initiating allopurinol.
ABSTRACT
ABSTRACT: Growing evidence from pharmacovigilance data and postmortem toxicology reports highlights the misuse potential of gabapentinoids. This study aimed to investigate the risk of serious adverse outcomes (drug misuse, overdose, major trauma), and their risk factors, in primary care patients who are prescribed gabapentinoids. Using the UK Clinical Practice Research Datalink, a matched cohort study calculated adverse event rates separately for gabapentinoid-exposed and unexposed cohorts. In the exposed cohort, event rates for exposure to a range of potential risk factors were calculated. Event rates were compared using Cox proportional hazards models, adjusted for age, sex, deprivation, previous mental health diagnosis, and coprescribing with potentially interacting medicines. Substance misuse (gabapentin adjusted hazard ratio [95% CI]: 2.40 [2.25-2.55]), overdose (2.99 [2.56-3.49]), and major trauma (0-2.5 years: 1.35 [1.28-1.42]; 2.5 to 10 years: 1.73 [1.56-1.95]) were more common among patients prescribed gabapentinoids than matched individuals who were not. The association with overdose was stronger for pregabalin than gabapentin. All adverse outcomes were significantly associated with smoking, history of substance misuse, overdose, or a mental health condition and prescription of opioids, benzodiazepines, antidepressants, and Z-drug hypnotics (eg, gabapentin hazard ratios for association of concurrent opioid use: misuse 1.49 [1.47-1.51]; overdose 1.87 [1.78-1.96]; major trauma 1.28 [1.26-1.30]). Our findings highlight the importance of careful patient selection when prescribing gabapentinoids and the need to educate prescribers about the risks of these drugs, particularly in combination with other central nervous system depressants.
ABSTRACT
BACKGROUND AND OBJECTIVE: Protocols are invaluable documents for any research study, especially for prediction model studies. However, the mere existence of a protocol is insufficient if key details are omitted. We reviewed the reporting content and details of the proposed design and methods reported in published protocols for prediction model research. METHODS: We searched MEDLINE, Embase, and the Web of Science Core Collection for protocols for studies developing or validating a diagnostic or prognostic model using any modeling approach in any clinical area. We screened protocols published between Jan 1, 2022 and June 30, 2022. We used the abstract, introduction, methods, and discussion sections of The Transparent Reporting of a multivariable prediction model of Individual Prognosis Or Diagnosis (TRIPOD) statement to inform data extraction. RESULTS: We identified 30 protocols, of which 28 were describing plans for model development and six for model validation. All protocols were open access, including a preprint. 15 protocols reported prospectively collecting data. 21 protocols planned to use clustered data, of which one-third planned methods to account for it. A planned sample size was reported for 93% development and 67% validation analyses. 16 protocols reported details of study registration, but all protocols reported a statement on ethics approval. Plans for data sharing were reported in 13 protocols. CONCLUSION: Protocols for prediction model studies are uncommon, and few are made publicly available. Those that are available were reasonably well-reported and often described their methods following current prediction model research recommendations, likely leading to better reporting and methods in the actual study.
Subject(s)
Guideline Adherence , Humans , Guideline Adherence/statistics & numerical data , Research Design/standards , Models, StatisticalABSTRACT
This paper begins with the common phrase 'good girl' as a lens through which to explore the insidious nature of patronising and paternalistic language on women's agency in obstetric care. Here we see how misogynistic language is both violence against women in its own right, and serves to create a context in which more extreme obstetric violence can be precipitated. Based on thematic analysis of discussion on Mumsnet, and on contributions to a research-focused Facebook group, this paper illustrates the complexity of recognising and refuting misogyny as a female patient as well as the damage that can occur from a cultural context in which this language is normalised. Here, words both boast a materiality through the environments they reify, and become transient and slippery, with semiotic uncertainty.
Subject(s)
Death , Physicians , Pregnancy , Female , Humans , Fertility , Focus Groups , United KingdomABSTRACT
OBJECTIVE: To describe the frequency of open science practices in a contemporary sample of studies developing prognostic models using machine learning methods in the field of oncology. STUDY DESIGN AND SETTING: We conducted a systematic review, searching the MEDLINE database between December 1, 2022, and December 31, 2022, for studies developing a multivariable prognostic model using machine learning methods (as defined by the authors) in oncology. Two authors independently screened records and extracted open science practices. RESULTS: We identified 46 publications describing the development of a multivariable prognostic model. The adoption of open science principles was poor. Only one study reported availability of a study protocol, and only one study was registered. Funding statements and conflicts of interest statements were common. Thirty-five studies (76%) provided data sharing statements, with 21 (46%) indicating data were available on request to the authors and seven declaring data sharing was not applicable. Two studies (4%) shared data. Only 12 studies (26%) provided code sharing statements, including 2 (4%) that indicated the code was available on request to the authors. Only 11 studies (24%) provided sufficient information to allow their model to be used in practice. The use of reporting guidelines was rare: eight studies (18%) mentioning using a reporting guideline, with 4 (10%) using the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis Or Diagnosis statement, 1 (2%) using Minimum Information About Clinical Artificial Intelligence Modeling and Consolidated Standards Of Reporting Trials-Artificial Intelligence, 1 (2%) using Strengthening The Reporting Of Observational Studies In Epidemiology, 1 (2%) using Standards for Reporting Diagnostic Accuracy Studies, and 1 (2%) using Transparent Reporting of Evaluations with Nonrandomized Designs. CONCLUSION: The adoption of open science principles in oncology studies developing prognostic models using machine learning methods is poor. Guidance and an increased awareness of benefits and best practices of open science are needed for prediction research in oncology.
Subject(s)
Artificial Intelligence , Machine Learning , Humans , PrognosisABSTRACT
OBJECTIVES: Despite little evidence that analgesics are effective in inflammatory arthritis (IA), studies report substantial opioid prescribing. The extent this applies to other analgesics is uncertain. We undertook a comprehensive evaluation of analgesic prescribing in patients with IA in the Clinical Practice Research Datalink Aurum to evaluate this. METHODS: From 2004 to 2020, cross-sectional analyses evaluated analgesic prescription annual prevalence in RA, PsA and axial spondyloarthritis (axSpA), stratified by age, sex, ethnicity, deprivation and geography. Joinpoint regression evaluated temporal prescribing trends. Cohort studies determined prognostic factors at diagnosis for chronic analgesic prescriptions using Cox proportional hazards models. RESULTS: Analgesic prescribing declined over time but remained common: 2004 and 2020 IA prescription prevalence was 84.2/100 person-years (PY) (95% CI 83.9, 84.5) and 64.5/100 PY (64.2, 64.8), respectively. In 2004, NSAIDs were most prescribed (56.1/100 PY; 55.8, 56.5), falling over time. Opioids were most prescribed in 2020 (39.0/100 PY; 38.7, 39.2). Gabapentinoid prescribing increased: 2004 prevalence 1.1/100 PY (1.0, 1.2); 2020 prevalence 9.9/100 PY (9.7, 10.0). Most opioid prescriptions were chronic (2020 prevalence 23.4/100 PY [23.2, 23.6]). Non-NSAID analgesic prescribing was commoner in RA, older people, females and deprived areas/northern England. Conversely, NSAID prescribing was commoner in axSpA/males, varying little by deprivation/geography. Peri-diagnosis was high-risk for starting chronic opioid/NSAID prescriptions. Prognostic factors for chronic opioid/gabapentinoid and NSAID prescriptions differed, with NSAIDs having no consistently significant association with deprivation (unlike opioids/gabapentinoids). CONCLUSION: IA analgesic prescribing of all classes is widespread. This is neither evidence-based nor in line with guidelines. Peri-diagnosis is an opportune moment to reduce chronic analgesic prescribing.
ABSTRACT
OBJECTIVES: To determine the risk of adverse events associated with colchicine or non-steroidal anti-inflammatory drug (NSAID) prophylaxis when initiating allopurinol for gout. METHODS: We conducted two matched retrospective cohort studies in linked UK Clinical Practice Research Datalink and Hospital Episode Statistics datasets. Adults initiating allopurinol for gout with (1) colchicine or (2) NSAID prophylaxis were compared with those initiating without prophylaxis, individually matched by age, sex and propensity to receive the relevant prophylaxis. Weighted Cox proportional hazards models investigated associations between colchicine/NSAID and specified adverse events. RESULTS: 13 945 individuals prescribed colchicine were matched to 13 945 with no prophylaxis and 25 980 prescribed NSAID to 25 980 with no prophylaxis. Adverse event incidence rates were <200/10 000 patient-years except diarrhoea (784.4; 95% CI 694.0 to 886.5) and nausea (208.1; 95% CI 165.4 to 261.7) for colchicine and angina for NSAID (466.6; 95% CI 417.2 to 521.8). Diarrhoea (HR 2.22; 95% CI 1.83 to 2.69), myocardial infarction (MI) (1.55; 95% CI 1.10, 2.17), neuropathy (4.75; 95% CI 1.20 to 18.76), myalgia (2.64; 95% CI 1.45 to 4.81), bone marrow suppression (3.29; 95% CI 1.43 to 7.58) and any adverse event (1.91, 95% CI 1.65 to 2.20) were more common with colchicine than no prophylaxis, but not nausea/vomiting (1.34; 95% CI 0.97 to 1.85). Angina (1.60; 95% CI 1.37 to 1.86), acute kidney injury (1.56; 95% CI 1.20 to 2.03), MI (1.89; 95% CI 1.44 to 2.48), peptic ulcer disease (1.67; 95% CI 1.14 to 2.44) and any adverse event (1.63; 95% CI 1.44 to 1.85) were more common with NSAID than without. CONCLUSIONS: Adverse events were more common when allopurinol was initiated with prophylaxis, particularly diarrhoea with colchicine. Other events were uncommon, providing reassurance for patients and clinicians to enable shared decision-making.
Subject(s)
Gout , Myocardial Infarction , Adult , Humans , Colchicine/adverse effects , Allopurinol/adverse effects , Uric Acid , Gout Suppressants/adverse effects , Retrospective Studies , Propensity Score , Gout/drug therapy , Anti-Inflammatory Agents, Non-Steroidal/adverse effects , Cohort Studies , Myocardial Infarction/chemically induced , Diarrhea/chemically induced , Diarrhea/epidemiology , Diarrhea/prevention & control , United Kingdom/epidemiologyABSTRACT
Before embarking on an individual participant data meta-analysis (IPDMA) project, researchers should consider the power of their planned IPDMA conditional on the studies promising their IPD and their characteristics. Such power estimates help inform whether the IPDMA project is worth the time and funding investment, before IPD are collected. Here, we suggest how to estimate the power of a planned IPDMA of randomised trials aiming to examine treatment-covariate interactions at the participant-level (i.e., treatment effect modifiers). We focus on a time-to-event (survival) outcome with a binary or continuous covariate, and propose an approximate analytic power calculation that conditions on the actual characteristics of trials, for example, in terms of sample sizes and covariate distributions. The proposed method has five steps: (i) extracting the following aggregate data for each group in each trial-the number of participants and events, the mean and SD for each continuous covariate, and the proportion of participants in each category for each binary covariate; (ii) specifying a minimally important interaction size; (iii) deriving an approximate estimate of Fisher's information matrix for each trial and the corresponding variance of the interaction estimate per trial, based on assuming an exponential survival distribution; (iv) deriving the estimated variance of the summary interaction estimate from the planned IPDMA, under a common-effect assumption, and (v) calculating the power of the IPDMA based on a two-sided Wald test. Stata and R code are provided and a real example provided for illustration. Further evaluation in real examples and simulations is needed.
Subject(s)
Sample Size , Humans , Randomized Controlled Trials as TopicABSTRACT
Background: The UK government reclassified gabapentin and pregabalin as 'controlled drugs' from April 2019. This study aimed to describe the trends in gabapentinoid prescribing before and immediately after reclassification, in the UK Clinical Practice Research Datalink, an electronic primary care health record broadly representative of the UK. Methods: Separately for gabapentin and pregabalin, we calculated annual incident and prevalent prescribing rates from year of UK approval (April 1997 and 2004 respectively) to September 2019, and monthly incident and prevalent prescribing rates (October 2017-September 2019). Significant changes in temporal trends were determined using joinpoint regression. We also described potential prescribing indications, prior pain-related prescribing, and co-prescribing with potentially interacting medicines. Findings: Incident gabapentin prescribing increased annually, peaking in 2016-17, at 625/100,000 patient years before falling steadily to 2019. Incident pregabalin prescribing peaked at 329/100,000 patient years in 2017-18 and did not fall significantly until 2019. Prevalent gabapentin and pregabalin prescribing increased annually to 2017-18 and 2018-19 respectively, before plateauing. Gabapentinoids were commonly co-prescribed with opioids (60%), antidepressants (52%), benzodiazepines (19%), and Z-drugs (10%). Interpretation: Following a dramatic rise, incident gabapentinoid prescribing has started to fall but the specific impact of reclassification on prescribing rates remains unclear. Limited change in prevalent gabapentinoid prescribing during the 6 months following their reclassification as controlled drugs suggests little immediate impact on continued gabapentinoid prescribing for existing users. Funding: National Institute for Health and Care Research (NIHR) Research for Patient Benefit Programme. NIHR Applied Research Collaboration West Midlands. NIHR School for Primary Care Research.
ABSTRACT
INTRODUCTION: The aim of this study was to generate a baseline database of print media reporting on alcohol and other drug (AOD) issues prior to the release of the Mindframe guidelines in March 2019. Specifically, to: (i) describe the content associated with media entries that focus on AOD use in Australian news media; (ii) determine how the media entries compare to several domains associated with recently developed Mindframe guidelines for publicly reporting on AOD; and (iii) identify content factors associated with different scores. METHODS: Media entries between July 2016 and June 2017 were searched for key AOD-related terms using the Australian and New Zealand Newsstream database. Two coding schemes were developed to rate a stratified sample of 50% of the media entries against the Mindframe guidelines. Associations between content and total comparison scores were determined using linear regression models. RESULTS: Detailed coding of the 2007 articles identified as relevant for the current study indicated that a majority (67%) were focused on one of three substances: alcohol, cannabis or methamphetamine. Most of the entries were either law enforcement (22%) or criminal justice related (19%). Entries that focused on methamphetamine scored significantly lower than entries on alcohol when compared to the Mindframe guidelines, similarly entries focused on crime/justice-related topics scored significantly lower than entries focused on positive outcomes. DISCUSSION AND CONCLUSIONS: A disproportionate number of print media entries, particularly those related methamphetamine use, focused on crime or justice-related topics, potentially further contributing to stigma, and emphasising the legal consequences of AOD use.
Subject(s)
Alcohol-Related Disorders , Methamphetamine , Substance-Related Disorders , Humans , Australia/epidemiology , Substance-Related Disorders/epidemiology , Mass Media , EthanolABSTRACT
Background: Contemporary data on rheumatoid arthritis (RA), psoriatic arthritis (PsA) and axial spondyloarthritits (SpA) epidemiology in England are lacking. This knowledge is crucial to planning healthcare services. We updated algorithms defining patients with diagnoses of RA, PsA, and axial SpA in primary care and applied them to describe their incidence and prevalence in the Clinical Practice Research Datalink Aurum, an electronic health record (EHR) database covering â¼20% of England. Methods: Algorithms for ascertaining patients with RA, axial SpA, and PsA diagnoses validated in primary care EHR databases using Read codes were updated (to account for the English NHS change to SNOMED CT diagnosis coding) and applied. Updated diagnosis and synthetic disease-modifying anti-rheumatic drug code lists were devised by rheumatologists and general practitioners. Annual incidence/point-prevalence of RA, PsA, and axial SpA diagnoses were calculated from 2004 to 2020 and stratified by age/sex. Findings: Point-prevalence of RA/PsA diagnoses increased annually, peaking in 2019 (RA 0·779% [95% confidence interval (CI) 0·773, 0·784]; PsA 0·287% [95% CI 0·284, 0·291]) then falling slightly. Point-prevalence of axial SpA diagnoses increased annually (except in 2018/2019), peaking in 2020 (0·113% [95% CI 0·111, 0·115]). RA diagnosis annual incidence was higher between 2013-2019 (after inclusion in the Quality and Outcomes Framework, range 49·1 [95% CI 47·7, 50·5] to 52·1 [95% CI 50·6, 53·6]/100,000 person-years) than 2004-2012 (range 34·5 [95% CI 33·2, 35·7] to 40·0 [95% CI 38·6, 41·4]/100,000 person-years). Increases in the annual incidence of PsA/axial SpA diagnosis occurred following new classification criteria publication. Annual incidence of RA, PsA and axial SpA diagnoses fell by 40·1%, 67·4%, and 38·1%, respectively between 2019 and 2020, likely reflecting the COVID-19 pandemic's impact on their diagnosis. Interpretation: Recorded RA, PsA, and axial SpA diagnoses are increasingly prevalent in England, underlining the importance of organising healthcare services to provide timely, treat-to-target care to optimise the health of >1% of adults in England. Funding: National Institute for Health and Care Research (NIHR300826).
ABSTRACT
Before embarking on an individual participant data meta-analysis (IPDMA) project, researchers and funders need assurance it is worth their time and cost. This should include consideration of how many studies are promising their IPD and, given the characteristics of these studies, the power of an IPDMA including them. Here, we show how to estimate the power of a planned IPDMA of randomized trials to examine treatment-covariate interactions at the participant level (ie, treatment effect modifiers). We focus on a binary outcome with binary or continuous covariates, and propose a three-step approach, which assumes the true interaction size is common to all trials. In step one, the user must specify a minimally important interaction size and, for each trial separately (eg, as obtained from trial publications), the following aggregate data: the number of participants and events in control and treatment groups, the mean and SD for each continuous covariate, and the proportion of participants in each category for each binary covariate. This allows the variance of the interaction estimate to be calculated for each trial, using an analytic solution for Fisher's information matrix from a logistic regression model. Step 2 calculates the variance of the summary interaction estimate from the planned IPDMA (equal to the inverse of the sum of the inverse trial variances from step 1), and step 3 calculates the corresponding power based on a two-sided Wald test. Stata and R code are provided, and two examples given for illustration. Extension to allow for between-study heterogeneity is also considered.
Subject(s)
Data Analysis , Models, Statistical , Humans , Randomized Controlled Trials as TopicABSTRACT
BACKGROUND: The prevention of self-harm is an international public health priority. It is vital to identify at-risk populations, particularly as self-harm is a risk factor for suicide. This study aims to examine the risk of self-harm in people with vertebral fractures. METHODS: Retrospective cohort study. Patients with vertebral fracture were identified within the Clinical Practice Research Datalink and matched to patients without fracture by sex and age. Incident self-harm was defined by primary care record codes following vertebral fracture. Overall incidence rates (per 10,000 person-years (PY)) were reported. Cox regression analysis determined risk (hazard ratios (HR), 95 % confidence interval (CI)) of self-harm compared to the matched unexposed cohort. Initial crude analysis was subsequently adjusted and stratified by median age and sex. RESULTS: The number of cases of vertebral fracture was 16,293, with a matched unexposed cohort of the same size. Patients were predominantly female (70.1 %), median age was 76.3 years. Overall incidence of self-harm in the cohort with vertebral fracture was 12.2 (10.1, 14.8) /10,000 PY. There was an initial crude association between vertebral fracture and self-harm, which remained after adjustment (HR 2.4 (95 %CI 1.5, 3.6). Greatest risk of self-harm was found in those with vertebral fractures who were aged below 76.3 years (3.2(1.8, 5.7)) and male (3.9(1.8, 8.5)). CONCLUSIONS: Primary care patients with vertebral fracture are at increased risk of self-harm compared to people without these fractures. Male patients aged below 76 years of age appear to be at greatest risk of self-harm. Clinicians need to be aware of the potential for self-harm in this patient group.
Subject(s)
Self-Injurious Behavior , Spinal Fractures , Aged , Cohort Studies , Female , Humans , Incidence , Male , Retrospective Studies , Risk Factors , Self-Injurious Behavior/diagnosis , Self-Injurious Behavior/epidemiology , Spinal Fractures/epidemiologyABSTRACT
OBJECTIVE: To examine the risk of self-harm in rheumatic conditions. METHODS: We conducted a retrospective cohort study using data from the Clinical Practice Research Datalink. Patients with ankylosing spondylitis, fibromyalgia, osteoarthritis, or rheumatoid arthritis were identified from 1990 to 2016 and matched to patients without these conditions. Incident self-harm was defined by medical record codes following a rheumatic diagnosis. Incidence rates (per 10,000 person-years) were reported for each condition, both overall and year-on-year (2000-2016). Cox regression analysis determined risk (hazard ratio [HR] and 95% confidence interval [95% CI]) of self-harm for each rheumatic cohort compared to the matched unexposed cohort. Initial crude analysis was subsequently adjusted and stratified by age and sex. Due to nonproportionality over time, osteoarthritis was also stratified by disease duration (<1 year, ≥1 to <5 years, ≥5 to <10 years, and ≥10 years). RESULTS: The incidence of self-harm was highest in patients with fibromyalgia (HR 25.12 [95% CI 22.45-28.11] per 10,000 person-years) and lowest for osteoarthritis (HR 6.48 [95% CI 6.20-6.76]). There was a crude association with each rheumatic condition and self-harm, except for ankylosing spondylitis. Although attenuated, these associations remained after adjustment for fibromyalgia (HR 2.06 [95% CI 1.60-2.65]), rheumatoid arthritis (HR 1.59 [95% CI 1.20-2.11]), and osteoarthritis (1 to <5 years HR 1.12 [95% CI 1.01-1.24]; ≥5 to <10 years HR 1.35 [95% CI 1.18-1.54]). Age and sex were weak effect modifiers for these associations. CONCLUSION: Primary care patients with fibromyalgia, osteoarthritis, or rheumatoid arthritis (but not ankylosing spondylitis) are at increased risk of self-harm compared to people without these rheumatic conditions. Clinicians need to be aware of the potential for self-harm in patients with rheumatic conditions (particularly fibromyalgia), explore mood and risk with them, and offer appropriate support and management.
Subject(s)
Rheumatic Diseases/complications , Self-Injurious Behavior/etiology , Adult , Aged , Arthritis, Rheumatoid/complications , Arthritis, Rheumatoid/diagnosis , Arthritis, Rheumatoid/psychology , Databases, Factual , Female , Fibromyalgia/complications , Fibromyalgia/diagnosis , Fibromyalgia/psychology , Humans , Incidence , Male , Middle Aged , Osteoarthritis/complications , Osteoarthritis/diagnosis , Osteoarthritis/psychology , Retrospective Studies , Rheumatic Diseases/diagnosis , Rheumatic Diseases/psychology , Risk Assessment , Risk Factors , Self-Injurious Behavior/diagnosis , Self-Injurious Behavior/psychology , Time FactorsABSTRACT
OBJECTIVES: When developing a clinical prediction model, penalization techniques are recommended to address overfitting, as they shrink predictor effect estimates toward the null and reduce mean-square prediction error in new individuals. However, shrinkage and penalty terms ('tuning parameters') are estimated with uncertainty from the development data set. We examined the magnitude of this uncertainty and the subsequent impact on prediction model performance. STUDY DESIGN AND SETTING: This study comprises applied examples and a simulation study of the following methods: uniform shrinkage (estimated via a closed-form solution or bootstrapping), ridge regression, the lasso, and elastic net. RESULTS: In a particular model development data set, penalization methods can be unreliable because tuning parameters are estimated with large uncertainty. This is of most concern when development data sets have a small effective sample size and the model's Cox-Snell R2 is low. The problem can lead to considerable miscalibration of model predictions in new individuals. CONCLUSION: Penalization methods are not a 'carte blanche'; they do not guarantee a reliable prediction model is developed. They are more unreliable when needed most (i.e., when overfitting may be large). We recommend they are best applied with large effective sample sizes, as identified from recent sample size calculations that aim to minimize the potential for model overfitting and precisely estimate key parameters.
Subject(s)
Epidemiologic Research Design , Models, Statistical , Uncertainty , Bias , Humans , Prognosis , Reproducibility of Results , Sample SizeABSTRACT
This article reflects on insights from an action research project where we worked with students whose university experience was inhibited by the fear of failure. In contrast to the popular concept of 'learning from failure', which involves intellectualizing the experience and distancing ourselves from it, our findings demonstrate the importance of a 'present tense' focus on emotions and affects in order to understand the experience of failure for students. Doing so brings us face-to-face with the often painful experience of failure in the present moment which, we argue, is an important and valid part of the university experience. We conclude by reflecting on the kinds of spaces and skills that may be needed to work with this new understanding of failure and show that developing these is a crucial part of resisting neoliberalism and creating a more 'care-full' (Mountz et al., 2015) academy.
ABSTRACT
OBJECTIVES: Despite many shared risk factors and pathophysiological pathways, the risk of ischaemic heart disease (IHD) and myocardial infarction (MI) in interstitial lung disease (ILD) remains poorly understood. This lack of data could be preventing patients who may benefit from screening for these cardiovascular diseases from receiving it. METHODS: A population-based cohort study used electronic patient records from the Clinical Practice Research Datalink and linked Hospital Episode Statistics to identify 68 572 patients (11 688 ILD exposed (mean follow-up: 3.8 years); 56 884 unexposed controls (mean follow-up: 4.0 years), with 349 067 person-years of follow-up. ILD-exposed patients (pulmonary sarcoidosis (PS) or idiopathic pulmonary fibrosis (PF)) were matched (by age, sex, registered general practice and available follow-up time) to patients without ILD or IHD/MI. Rates of incident MI and IHD were estimated. HRs were modelled using multivariable Cox proportional hazards regression accounting for potential confounders. RESULTS: ILD was independently associated with IHD (HR 1.85, 95% CI 1.56 to 2.18) and MI (HR 1.74, 95% CI 1.44 to 2.11). In all disease categories, risk of both IHD and MI peaked between ages 60 and 69 years, except for the risk of MI in PS which was greatest <50 years. Men with PF were at greatest risk of IHD, while women with PF were at greatest risk of MI. CONCLUSIONS: ILD, particularly PF, is independently associated with MI and IHD after adjustment for established cardiovascular risk factors. Our results suggest clinicians should prioritise targeted assessment of cardiovascular risk in patients with ILD, particularly those aged 60-69 years. Further research is needed to understand the impact of such an approach to risk management.
Subject(s)
Lung Diseases, Interstitial/epidemiology , Myocardial Infarction/epidemiology , Myocardial Ischemia/epidemiology , Adult , Age Distribution , Age Factors , Aged , Aged, 80 and over , Case-Control Studies , Databases, Factual , Female , Heart Disease Risk Factors , Humans , Incidence , Lung Diseases, Interstitial/diagnosis , Male , Middle Aged , Myocardial Infarction/diagnosis , Myocardial Ischemia/diagnosis , Prognosis , Risk Assessment , Sex Distribution , Sex Factors , Time Factors , United Kingdom/epidemiologyABSTRACT
OBJECTIVES: To investigate potential subgroups of primary care-diagnosed patients with PMR based on self-reported pain and stiffness severity over time. METHODS: A total of 652 people with an incident PMR diagnosis were recruited from English general practices and completed a baseline postal questionnaire. They were followed up with a further six questionnaires over a 2 year period. A total of 446 people completed the 2 year follow-up. Pain and stiffness were reported on a 0-10 numerical rating scale. Latent class growth analysis was used to estimate the joint trajectories of pain and stiffness over time. A combination of statistical and clinical considerations was used to choose the number of clusters. Characteristics of the classes were described. RESULTS: Five clusters were identified. One cluster represented the profile of 'classical' PMR symptoms and one represented sustained symptoms that may not be PMR. The other three clusters displayed a partial recovery, a recovery followed by worsening and a slow, but sustained recovery. Those displaying classical PMR symptoms were in better overall health at diagnosis than the other groups. CONCLUSION: PMR is a heterogeneous condition, with a number of phenotypes. The spectrum of presentation, as well as varying responses to treatment, may be related to underlying health status at diagnosis. Future research should seek to stratify patients at diagnosis to identify those likely to have a poor recovery and in need of an alternative treatment pathway. Clinicians should be aware of the different experiences of patients and monitor symptoms closely, even where there is initial improvement.
Subject(s)
Pain/etiology , Polymyalgia Rheumatica/complications , Aged , Aged, 80 and over , Female , Humans , Male , Middle Aged , Prognosis , Surveys and QuestionnairesABSTRACT
BACKGROUND: The objectives of this study were to estimate the population prevalence and distribution of plantar heel pain in mid-to-older age groups, examine associations with selected health status and lifestyle factors, and report the frequency of healthcare use. METHODS: Adults aged ≥50 years registered with four general practices were mailed a health survey (n = 5109 responders). Plantar heel pain in the last month was defined by self-reported shading on a foot manikin, and was defined as disabling if at least one of the function items of the Manchester Foot Pain and Disability Index were also reported. Population prevalence estimates and associations between plantar heel pain and demographic characteristics, health status measures and lifestyle factors were estimated using multiple imputation and weighted logistic regression. Healthcare professional consultation was summarised as the 12-month period prevalence of foot pain-related consultation. RESULTS: The population prevalence of plantar heel pain was 9.6% (95% CI: 8.8, 10.5) and 7.9% (7.1, 8.7) for disabling plantar heel pain. Occurrence was slightly higher in females, comparable across age-groups, and significantly higher in those with intermediate/routine and manual occupations. Plantar heel pain was associated with physical and mental impairment, more anxiety and depression, being overweight, a low previous use of high-heeled footwear, and lower levels of physical activity and participation. The 12-month period prevalence of foot pain-related consultation with a general practitioner, physiotherapist or podiatrist/chiropodist was 43.0, 15.1 and 32.8%, respectively. CONCLUSIONS: Plantar heel pain is a common, disabling symptom among adults aged 50 years and over. Observed patterns of association indicate that in addition to focused foot-specific management, primary care interventions should also target more general physical and psychological factors that could potentially act as barriers to treatment adherence and recovery.