Development of a clinical risk score for pain and function following total knee arthroplasty: results from the TRIO study

Influencing factors on the quality of recovery after total knee arthroplasty: development of a predictive model

Introduction

Total Knee Arthroplasty (TKA) is a widely performed procedure that significantly benefits patients with severe knee degeneration. However, the recovery outcomes post-surgery can vary significantly among patients. Identifying the factors influencing these outcomes is crucial for improving patient care and satisfaction.

Methods

In this retrospective study, we analyzed 362 TKA cases performed between January 1, 2018, and July 1, 2022. Multivariate logistic regression was employed to identify key predictors of recovery within the first year after surgery.

Results

The analysis revealed that Body Mass Index (BMI), age-adjusted Charlson Comorbidity Index (aCCI), sleep quality, Bone Mineral Density (BMD), and analgesic efficacy were significant predictors of poor recovery (p < 0.05). These predictors were used to develop a clinical prediction model, which demonstrated strong predictive ability with an Area Under the Receiver Operating Characteristic (AUC) curve of 0.802. The model was internally validated.

Discussion

The findings suggest that personalized postoperative care and tailored rehabilitation programs based on these predictors could enhance recovery outcomes and increase patient satisfaction following TKA.

Risk factors of chronic postoperative pain after total knee arthroplasty: a systematic review

BACKGROUND

There is a lack of relevant studies to grade the evidence on the risk factors of chronic pain after total knee arthroplasty (TKA), and only quantitative methods are used for systematic evaluation. The review aimed to systematically identify risk factors of chronic postoperative pain following TKA and to evaluate the strength of the evidence underlying these correlations.

METHODS

PubMed, Web of Science, Cochrane Library, Embase, and CINAHL databases were searched from initiation to September 2023. Cohort studies, case-control studies, and cross-sectional studies involving patients undergoing total knee replacement were included. A semi-quantitative approach was used to grade the strength of the evidence-based on the number of investigations, the quality of the studies, and the consistency of the associations reported by the studies.

RESULTS

Thirty-two articles involving 18,792 patients were included in the final systematic review. Ten variables were found to be strongly associated with postoperative pain, including Age, body mass index (BMI), comorbidities condition, preoperative pain, chronic widespread pain, preoperative adverse health beliefs, preoperative sleep disorders, central sensitization, preoperative anxiety, and preoperative function. Sixteen factors were identified as inconclusive evidence.

CONCLUSIONS

This systematic review clarifies which risk factors could be involved in future research on TKA pain management for surgeons and patients. It highlights those factors that have been controversial or weakly correlated, emphasizing the need for further high-quality studies to validate them. Most crucially, it can furnish clinicians with vital information regarding high-risk patients and their clinical attributes, thereby aiding in the development of preventive strategies to mitigate postoperative pain following TKA.

TRIAL REGISTRATION

This systematic review has been registered on the PROSPERO platform (CRD42023444097).

A mapping review on preoperative prognostic factors and outcome measures of revision total knee arthroplasty

To map literature on prognostic factors related to outcomes of revision total knee arthroplasty (rTKA), to identify extensively studied factors and to guide future research into what domains need further exploration. We performed a systematic literature search in MEDLINE, Embase, and Web of Science. The search string included multiple synonyms of the following keywords: "revision TKA", "outcome" and "prognostic factor". We searched for studies assessing the association between at least one prognostic factor and at least one outcome measure after rTKA surgery. Data on sample size, study design, prognostic factors, outcomes, and the direction of the association was extracted and included in an evidence map. After screening of 5,660 articles, we included 166 studies reporting prognostic factors for outcomes after rTKA, with a median sample size of 319 patients (30 to 303,867). Overall, 50% of the studies reported prospectively collected data, and 61% of the studies were performed in a single centre. In some studies, multiple associations were reported; 180 different prognostic factors were reported in these studies. The three most frequently studied prognostic factors were reason for revision (213 times), sex (125 times), and BMI (117 times). Studies focusing on functional scores and patient-reported outcome measures as prognostic factor for the outcome after surgery were limited (n = 42). The studies reported 154 different outcomes. The most commonly reported outcomes after rTKA were: re-revision (155 times), readmission (88 times), and reinfection (85 times). Only five studies included costs as outcome. Outcomes and prognostic factors that are routinely registered as part of clinical practice (e.g. BMI, sex, complications) or in (inter)national registries are studied frequently. Studies on prognostic factors, such as functional and sociodemographic status, and outcomes as healthcare costs, cognitive and mental function, and psychosocial impact are scarce, while they have been shown to be important for patients with osteoarthritis.

Instruments to assess appropriateness of hip and knee arthroplasty: a systematic review

OBJECTIVE

To assess criteria and psychometric properties of instruments for assessing appropriateness of elective joint arthroplasty (JA) for adults with primary hip and knee osteoarthritis (OA).

METHODS

A systematic review guided by Cochrane methods and PRISMA guidelines. Studies were searched in five databases. Eligible articles include all study designs developing, testing, and/or using an instrument to assess JA appropriateness. Two independent reviewers screened and extracted data. Instruments were compared with Hawker et al. JA consensus criteria. Psychometric properties of instruments were described and appraised guided by Fitzpatrick's and COSMIN approaches.

RESULTS

Of 55 instruments included, none met all Hawker et al. JA consensus criteria. Criteria the most met were pain (n = 50), function (n = 49), quality of life (n = 33), and radiography (n = 24). Criteria the least met were clinical evidence of OA (n = 18), expectations (n = 15), readiness for surgery (n = 11), conservative treatments (n = 8), and patient/surgeon agree benefits outweigh risks (n = 0). Instrument by Arden et al. met the most criteria (6 of 9). The most tested psychometric properties were appropriateness (n = 55), face/content validity (n = 55), predictive validity (n = 29), construct validity and feasibility (n = 24). The least tested psychometric properties were intra-rater reliability (n = 3), internal consistency (n = 5), and inter-rater reliability (n = 13). Instruments by Gutacker et al. and Osborne et al. met the most psychometric properties (4 of 10).

CONCLUSION

Most instruments included traditional criteria for assessing JA appropriateness but did not include a trial of conservative treatments or shared decision-making elements. There was limited evidence on psychometric properties.

Evaluating methodological quality of prognostic prediction models on patient reported outcome measurements after total hip replacement and total knee replacement surgery: a systematic review protocol

BACKGROUND

Prediction models for poor patient-reported surgical outcomes after total hip replacement (THR) and total knee replacement (TKR) may provide a method for improving appropriate surgical care for hip and knee osteoarthritis. There are concerns about methodological issues and the risk of bias of studies producing prediction models. A critical evaluation of the methodological quality of prediction modelling studies in THR and TKR is needed to ensure their clinical usefulness. This systematic review aims to (1) evaluate and report the quality of risk stratification and prediction modelling studies that predict patient-reported outcomes after THR and TKR; (2) identify areas of methodological deficit and provide recommendations for future research; and (3) synthesise the evidence on prediction models associated with post-operative patient-reported outcomes after THR and TKR surgeries.

METHODS

MEDLINE, EMBASE, and CINAHL electronic databases will be searched to identify relevant studies. Title and abstract and full-text screening will be performed by two independent reviewers. We will include (1) prediction model development studies without external validation; (2) prediction model development studies with external validation of independent data; (3) external model validation studies; and (4) studies updating a previously developed prediction model. Data extraction spreadsheets will be developed based on the CHARMS checklist and TRIPOD statement and piloted on two relevant studies. Study quality and risk of bias will be assessed using the PROBAST tool. Prediction models will be summarised qualitatively. Meta-analyses on the predictive performance of included models will be conducted if appropriate. A narrative review will be used to synthesis the evidence if there are insufficient data to perform meta-analyses.

DISCUSSION

This systematic review will evaluate the methodological quality and usefulness of prediction models for poor outcomes after THR or TKR. This information is essential to provide evidence-based healthcare for end-stage hip and knee osteoarthritis. Findings of this review will contribute to the identification of key areas for improvement in conducting prognostic research in this field and facilitate the progress in evidence-based tailored treatments for hip and knee osteoarthritis.

SYSTEMATIC REVIEW REGISTRATION

PROSPERO registration number CRD42021271828.

Preoperative Factors Affecting the Patient-Reported Outcome Measures following Total Knee Replacement: Socioeconomic Factors and Preoperative OKS Have a Clinically Meaningful Effect

The Oxford Knee Score (OKS) is a patient-reported outcome questionnaire typically used to assess function and pain in patients undergoing total knee replacement (TKR). However, research is inconclusive as to which preoperative factors are important in explaining variation in outcome following TKR. The operative records of 12,709 patients who underwent primary TKR over a 9-year period were analyzed. The following variables were collected for each patient: age, sex, body mass index (BMI), Index of Multiple Deprivation decile rank, side of operation, diagnosis, the American Society of Anaesthesiologists (ASA) grade, preoperative OKS, EQ-5D index score, EuroQol visual analog scale (EQ-VAS) score, the postoperative OKS at 1 and 2 years. Generalized linear regression models were performed at 1 and 2 years to investigate the effect of the preoperative variables on the postoperative OKS. The effect of age, sex, BMI, Index of Multiple Deprivation decile rank, diagnosis, ASA grade, preoperative OKS, EuroQoL five-dimensional (EQ-5D) index score, and EQ-VAS score were all statistically significant in explaining the variation in OKS at 1 and 2 years postoperatively, with critical level of significance of 0.05 (5%). Being male aged 60 to 69 years of normal BMI, ASA grade I (fit and healthy), living in an affluent area, not reporting preoperative anxiety/depression, were associated with an enhanced mean postoperative OKS at both 1 and 2 years. When adjusted for potential confounding, age of 60-69 years, male sex, normal BMI, lower ASA grade, higher Index of Multiple Deprivation and higher pre-operative EQ-5D, EQ-VAS and OKS were identified as factors that resulted in higher post-operative OKS after primary TKR.

Patient-Reported Outcomes for Function and Pain in Total Knee Arthroplasty Patients

BACKGROUND

Some patients undergoing total knee arthroplasty successfully manage their condition postoperatively, while others encounter challenges in regaining function and controlling pain during recovery at home.

OBJECTIVE

To use traditional statistics and machine learning to develop prediction models that identify patients likely to have increased care needs related to managing function and pain following total knee arthroplasty.

METHODS

This study included 201 patients. Outcomes were changes between baseline and follow-up in the functional and pain subcomponents of the Oxford Knee Score. Both classification and regression modeling were applied. Twenty-one predictors were included. Tenfold cross-validation was used, and the regression models were evaluated based on root mean square error, mean absolute error, and coefficient of determination. Classification models were evaluated based on the area under the receiver operating curve, sensitivity, and specificity.

RESULTS

In classification modeling, random forest and stochastic gradient boosting provided the best overall metrics for model performance. A support vector machine and a stochastic gradient boosting machine in regression modeling provided the best predictive performance. The models performed better in predicting challenges related to function compared to challenges related to pain.

DISCUSSION

There is valuable predictive information in the data routinely collected for patients undergoing total knee arthroplasty. The developed models may predict patients who are likely to have enhanced care needs regarding function and pain management. Improvements are needed before the models can be implemented in routine clinical practice.

Artificial intelligence in knee arthroplasty: current concept of the available clinical applications

Background

Artificial intelligence (AI) is defined as the study of algorithms that allow machines to reason and perform cognitive functions such as problem-solving, objects, images, word recognition, and decision-making. This study aimed to review the published articles and the comprehensive clinical relevance of AI-based tools used before, during, and after knee arthroplasty.

Methods

The search was conducted through PubMed, EMBASE, and MEDLINE databases from 2000 to 2021 using the 2009 Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocol (PRISMA).

Results

A total of 731 potential articles were reviewed, and 132 were included based on the inclusion criteria and exclusion criteria. Some steps of the knee arthroplasty procedure were assisted and improved by using AI-based tools. Before surgery, machine learning was used to aid surgeons in optimizing decision-making. During surgery, the robotic-assisted systems improved the accuracy of knee alignment, implant positioning, and ligamentous balance. After surgery, remote patient monitoring platforms helped to capture patients’ functional data.

Conclusion

In knee arthroplasty, the AI-based tools improve the decision-making process, surgical planning, accuracy, and repeatability of surgical procedures.

Patients follow three distinct outcome trajectories following total knee arthroplasty

AIMS

Although total knee arthroplasty (TKA) is a highly successful procedure, about 20% of patients remain dissatisfied postoperatively. This proportion is derived from dichotomous models of the assessment of surgical success or failure, which may not reflect the spectrum of outcomes. The aim of this study was to explore differing responses to surgery, and assess whether there are distinct groups of patients with differing patterns of outcome.

METHODS

This was a secondary analysis of a UK multicentre TKA longitudinal cohort study. We used a group-based trajectory modelling analysis of Oxford Knee Score (OKS) in the first year following surgery with longitudinal data involving five different timepoints and multiple predictor variables. Associations between the derived trajectory groups and categorical baseline variables were assessed, and predictors of trajectory group membership were identified using Poisson regression and multinomial logistic regression, as appropriate. The final model was adjusted for sociodemographic factors (age, sex) and baseline OKS.

RESULTS

Data from 731 patients were available for analysis. Three distinct trajectories of outcome were identified: "poor" 14.0%, "modest" 39.1%, and "good" 46.9%. The predicted probability of membership for patients assigned to each trajectory group was high (0.89 to 0.93). Preoperative mental, physical health, and psychosocial factors determined which trajectory is likely to be followed. Poor responders were characterized by a comparatively small number of factors, preoperative expectations of pain and limitations, coping strategies, and a lower baseline physical health status, while the good responders were characterized by a combination of clinical, psychosocial, mental health, and quality of life factors.

CONCLUSION

We identified three distinct response trajectories in patients undergoing TKA. Controlling for baseline score, age, and sex, psychosocial factors such as expectations of pain and limited function and poor coping strategies differentiated the trajectory groups, suggesting a role for preoperative psychosocial support in optimizing the clinical outcome. Cite this article: Bone Joint J 2021;103-B(6):1096-1102.

Prognostic models for knee osteoarthritis: a protocol for systematic review, critical appraisal, and meta-analysis

BACKGROUND

Osteoarthritis is the most common degenerative joint disease. It is associated with significant socioeconomic burden and poor quality of life, mainly due to knee osteoarthritis (KOA), and related total knee arthroplasty (TKA). Since early detection method and disease-modifying drug is lacking, the key of KOA treatment is shifting to disease prevention and progression slowing. The prognostic prediction models are called for to guide clinical decision-making. The aim of our review is to identify and characterize reported multivariable prognostic models for KOA about three clinical concerns: (1) the risk of developing KOA in the general population, (2) the risk of receiving TKA in KOA patients, and (3) the outcome of TKA in KOA patients who plan to receive TKA.

METHODS

The electronic datasets (PubMed, Embase, the Cochrane Library, Web of Science, Scopus, SportDiscus, and CINAHL) and gray literature sources (OpenGrey, British Library Inside, ProQuest Dissertations & Theses Global, and BIOSIS preview) will be searched from their inception onwards. Title and abstract screening and full-text review will be accomplished by two independent reviewers. The multivariable prognostic models that concern on (1) the risk of developing KOA in the general population, (2) the risk of receiving TKA in KOA patients, and (3) the outcome of TKA in KOA patients who plan to receive TKA will be included. Data extraction instrument and critical appraisal instrument will be developed before formal assessment and will be modified during a training phase in advance. Study reporting transparency, methodological quality, and risk of bias will be assessed according to the TRIPOD statement, CHARMS checklist, and PROBAST tool, respectively. Prognostic prediction models will be summarized qualitatively. Quantitative metrics on the predictive performance of these models will be synthesized with meta-analyses if appropriate.

DISCUSSION

Our systematic review will collate evidence from prognostic prediction models that can be used through the whole process of KOA. The review may identify models which are capable of allowing personalized preventative and therapeutic interventions to be precisely targeted at those individuals who are at the highest risk. To accomplish the prediction models to cross the translational gaps between an exploratory research method and a valued addition to precision medicine workflows, research recommendations relating to model development, validation, or impact assessment will be made.

SYSTEMATIC REVIEW REGISTRATION

PROSPERO CRD42020203543.

Clinical Decision Support Tools for Predicting Outcomes in Patients Undergoing Total Knee Arthroplasty: A Systematic Review

BACKGROUND

Total knee arthroplasty is the standard surgical treatment for end-stage osteoarthritis. Although widely accepted as a successful procedure, approximately 30% of patients are not satisfied due to non-optimal postoperative outcomes. Clinical decision support tools that are able to accurately predict post-surgery outcomes would assist in providing individualized advice or services to help alleviate possible issues, resulting in significant benefits to both the healthcare system and individuals.

METHODS

Five databases (Ovid Medline, Ovid EMBASE, CINAHL complete, Cochrane Library, and Scopus) were searched for the key phrases "knee replacement" or "knee arthroplasty" and "decision support tool," "decision tool," "predict∗ tool," "predict∗ model," "algorithm" or "nomogram." Searches were limited to peer-reviewed journal articles published between January 2000 and June 2019. Reference lists of included articles were examined. Authors came to a consensus on the final list of included articles.

RESULTS

Eighteen articles were included for review. Most models reported low predictive success and inability to externally validate. Both candidate and final predictor variables were inconsistent between studies. Only 1 model was considered strongly predictive (AUROC >0.8), and only 2 studies were able to externally validate their developed model. In general, models that performed well used large patient numbers, were tested on similar demographics, and used either nonlinear input transformations or a completely nonlinear model.

CONCLUSION

Some models do show promise; however, there remains the question of whether the reported predictive success can continue to be replicated. Furthermore, clinical applicability and interpretation of predictive tools should be considered during development.

Achieving Validated Thresholds for Clinically Meaningful Change on the Knee Injury and Osteoarthritis Outcome Score After Total Knee Arthroplasty: Findings From a University-based Orthopaedic Tertiary Care Safety Net Practice

A lack of knowledge exists about which patient characteristics predict failure to meet validated thresholds for clinically meaningful change on the Knee Injury and Osteoarthritis Outcome Score (KOOS) after total knee arthroplasty (TKA).

Methods

A retrospective chart review was performed on patients who underwent primary TKA by a single surgeon between January 2013 and June 2018. Variables included demographics (age, sex, race, and insurance type), comorbidities, body mass index, and preoperative KOOS subscale scores. Multivariate logistic regression was performed to identify characteristics associated with failing to meet or exceed the minimal clinically important difference (MCID) and substantial clinical benefit (SCB) on each KOOS subscale 6 months after TKA.

Results

A total of 159 patients were included. At 6 months after TKA, approximately one-third of patients (21% to 32%) failed to meet or exceed the MCID and 27% to 39% failed to meet or exceed the SCB on all KOOS subscales. Better preoperative KOOS Symptoms, quality of life, and activities of daily living subscale scores were statistically significantly associated with failing to meet the MCID and SCB on each respective subscale. Demographics, comorbidities, and body mass index were not notable predictors of either outcome for any of the KOOS subscales.

Discussion

About one-third of TKA patients in this single-site, single-surgeon sample failed to achieve a clinically meaningful outcome, and up to 4 in 10 patients had a less-than-ideal outcome 6 months after surgery. Surgeons should take care to set realistic expectations for patients with the least severe knee problems before TKA because this subgroup is especially at a high risk of failing to achieve a satisfactory outcome.

Effectiveness and reporting standards of psychological interventions for improving short-term and long-term pain outcomes after total knee replacement: a systematic review

OBJECTIVES

To assess the effectiveness and reporting standards of psychological interventions for improving outcomes after total knee replacement (TKR).

DESIGN

Medline, Embase, and PsycINFO were searched from inception to up to 9 May 2019 with no language restrictions applied. Randomised controlled trials (RCTs) assessing the effectiveness of psychological interventions for short-term and long-term postoperative pain after TKR were included. Screening, data extraction, and assessment of methodological quality were performed in duplicate by two reviewers. The primary effectiveness outcome was postoperative pain severity and the primary harm outcome was serious adverse events. Secondary outcomes included function, quality of life, and psychological well-being. Reporting standards were assessed using the Template for Intervention Description and Replication (TIDieR) checklist for intervention reporting.

RESULTS

12 RCTs were included, with a total of 1299 participants. Psychological interventions comprised music therapy (five studies), guided imagery and music (one study), hypnosis (one study), progressive muscle relaxation with biofeedback (one study), pain coping skills programme (one study), cognitive-behavioural therapy (two studies), and a postoperative management programme (one study). Due to the high heterogeneity of interventions and poor reporting of harms data, it was not possible to make any definitive statements about the overall effectiveness or safety of psychology interventions for pain outcomes after TKR.

CONCLUSION

Further evidence about the effectiveness of psychological interventions for improving pain outcomes after TKR is needed. The reporting of harm outcomes and intervention fidelity is currently poor and could be improved. Future work exploring the impact of intervention timing on effectiveness and whether different psychological approaches are needed to address acute postoperative pain and chronic postoperative pain would be of benefit.

PROSPERO REGISTRATION NUMBER

CRD42018095100.

[Predictors for persistent pain and dissatisfaction after total knee arthroplasty]

BACKGROUND

A substantial number of patients patients suffer from persistent pain or are unsatisfied after total knee arthroplasty (TKA).

OBJECTIVES

This work aims to present the frequency of postoperative persistent pain and/or dissatisfaction as well as known causes and predictors.

MATERIALS AND METHODS

The current literature is studied regarding the subject and is reviewed narratively.

RESULTS

Most postoperative problems did not arise from operation details, but from patient-related criteria, a lack of patient education and selection. The satisfaction correlates most strongly with the reduction of preoperative pain.

CONCLUSION

For a successful TKA, care should be taken that the following aspects are met preoperatively: clinically and radiologically advanced osteoarthritis, a patient age preferably older than 60 years, sufficient psychosocial resources to cope with postoperative stress, no opioid medication and realistic expectations after TKA. Postoperatively, patients with persistent pain or dissatisfaction should be checked for any prosthesis-related problems. If no prosthesis-related problems could be detected, the patients should be referred for interdisciplinary therapies.

Comparative responsiveness of the PROMIS-10 Global Health and EQ-5D questionnaires in patients undergoing total knee arthroplasty

AIMS

Responsiveness to clinically important change is a key feature of any outcome measure. Throughout Europe, health-related quality of life following total knee arthroplasty (TKA) is routinely measured with EuroQol five-dimension (EQ-5D) questionnaires. The Patient-Reported Outcomes Measurement Information System 10-Question Short-Form (PROMIS-10 Global Health) score is a new general heath outcome tool which is thought to offer greater responsiveness. Our aim was to compare these two tools.

PATIENTS AND METHODS

We accessed data from a prospective multicentre cohort study in the United Kingdom, which evaluated outcomes following TKA. The median age of the 721 patients was 69.0 years (interquartile range, 63.3 to 74.6). There was an even division of sex, and approximately half were educated to secondary school level. The preoperative EQ-5D, PROMIS-10, and Oxford Knee Scores (OKS) were available and at three, six, and 12 months postoperatively. Internal responsiveness was assessed by standardized response mean (SRM) and effect size (Cohen's d). External responsiveness was assessed by correlating change scores of the EQ-5D and PROMIS-10, with the minimal clinically important difference (MCID) of the OKS. Receiver operating characteristic (ROC) curves were used to assess the ability of change scores to discriminate between improved and non-improved patients.

RESULTS

All measures showed significant changes between the preoperative score and the various postoperative times (p < 0.001). Most improvement occurred during the first three months, with small but significant changes between three and six months, and no further change between six and 12 months postoperatively. SRM scores for EQ-5D, PROMIS-10, and OKS were large (> 0.8). ROC curves showed that both EQ-5D and PROMIS-10 were able to discriminate between patients who achieved the OKS MCID and those who did not (area under the curve (AUC) of 0.7 to 0.82).

CONCLUSION

The PROMIS-10 physical health tool showed greater responsiveness to change than the EQ-5D, most probably due to the additional questions on physical health parameters that are more susceptible to modification following TKA. The EQ-5D was, however, shown to be sensitive to clinically meaningful change following TKA, and provides the additional ability to calculate health economic utility scores. It is likely, therefore, that EQ-5D will continue to be the global health metric of choice in the United Kingdom. Cite this article: Bone Joint J 2019;101-B:832-837.

Optimising recruitment into trials using an internal pilot

Background

Recruitment to trials can be difficult. Despite careful planning and research that outlines ways to improve recruitment, many trials do not achieve their target on time and require extensions of funding or time.

Methods

We describe a trial in which an internal pilot with embedded qualitative research was used to improve recruitment processes and inform recruitment projections for the main trial. At the end of the pilot, it was clear that the sample size would not be met on time. Three steps were taken to optimise recruitment: (1) adjustments were made to the recruitment process using information from the qualitative work done in the pilot and advice from a patient and public involvement group, (2) additional recruiting sites were included based on site feasibility assessments and (3) a projection equation was used to estimate recruitment at each site and overall trial recruitment.

Results

Qualitative work during the pilot phase allowed us to develop strategies to optimise recruitment during the main trial, which were incorporated into patient information packs, the standard operating procedures and training sessions with recruiters. From our experience of feasibility assessments, we developed a checklist of recommended considerations for feasibility assessments. For recruitment projections, we developed a four-stage projection equation that estimates the number of participants recruited using a conversion rate of the number randomised divided by the number screened.

Conclusions

This work provides recommendations for feasibility assessments and an easy-to-use projection tool, which can be applied to other trials to help ensure they reach the required sample size.

Trial registration

ISRCTN, ISRCTN92545361. Registered on 6 September 2016.

Electronic supplementary material

The online version of this article (10.1186/s13063-019-3296-5) contains supplementary material, which is available to authorized users.

Classification using ensemble learning under weighted misclassification loss

Binary classification rules based on covariates typically depend on simple loss functions such as zero-one misclassification. Some cases may require more complex loss functions. For example, individual-level monitoring of HIV-infected individuals on antiretroviral therapy requires periodic assessment of treatment failure, defined as having a viral load (VL) value above a certain threshold. In some resource limited settings, VL tests may be limited by cost or technology, and diagnoses are based on other clinical markers. Depending on scenario, higher premium may be placed on avoiding false-positives, which brings greater cost and reduced treatment options. Here, the optimal rule is determined by minimizing a weighted misclassification loss/risk. We propose a method for finding and cross-validating optimal binary classification rules under weighted misclassification loss. We focus on rules comprising a prediction score and an associated threshold, where the score is derived using an ensemble learner. Simulations and examples show that our method, which derives the score and threshold jointly, more accurately estimates overall risk and has better operating characteristics compared with methods that derive the score first and the cutoff conditionally on the score especially for finite samples.