Kevin Kennedy
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Kevin Kennedy has a total of 133 co-authors. Here are the top co-authors by number of shared publications:
- Feng Xie
Department of Health Research Methods, Evidence, and Impact, McMaster University, Hamilton, Canada; Centre for Health Economics and Policy Analysis, McMaster University, Hamilton, Canada
5 shared publications - Jeffrey Andrew Sparks
Division of Rheumatology, Inflammation, and Immunity, Brigham and Women’s Hospital, Harvard Medical School, Boston, USA
3 shared publications - Julia Fridman Simard
Department of Epidemiology and Population Health, Stanford School of Medicine, Stanford, USA
3 shared publications Mitchell Levine
Department of Health Research Methods, Evidence and Impact (HEI), McMaster University Faculty of Health Sciences, Hamilton, Ontario, Canada
3 shared publications- Sebastian Eduardo Sattui
Division of Rheumatology and Clinical Immunology, Department of Medicine, University of Pittsburgh , Pittsburgh, PA
3 shared publications Tarin T Moni
Department of Biochemistry & Biomedical Sciences, McMaster University Faculty of Science, Hamilton, Ontario, Canada
3 shared publications
Latest Publications
- 27 February 2024
- Published by Springer Nature in The European Journal of Health Economics
- Vol. 25 (8) , 1417-1435
Systematic methodological review of health state values in glaucoma cost-utility analyses
Importance Describing the characteristics and sources of health state utility values and reporting practice in the literature of cost-utility analyses facilitates an understanding of the level of the transparency, validity, and generalizability of cost-utility analyses. Improving the quality of reporting will support investigators in describing the incremental value of emerging glaucoma interventions. Objective To describe the state of practice among published glaucoma cost-utility analysis studies, focusing on valuation of health and the quality of reporting. Evidence review We searched several databases including Medline, CINHAL, Embase, Web of Science, Scopus, Biosis previews, the Health Economic Evaluations Database, and the NHS Economic Evaluation Database (NHS EED). We included full-text, English, published cost-utility analyses of glaucoma interventions with quality-adjusted life years (QALYs) as the primary outcome measure to calculate incremental cost-utility ratios. Excluded studies were non-English language, reviews, editorials, protocols, or other types of economic studies (cost–benefit, cost-minimization, cost-effectiveness). Study characteristics, operational definitions of glaucoma health states and health state utilities were extracted. The original source of the health utility was reviewed to determine the scale of measurement and the source of preference weighting. Items from the Systematic Review of Utilities for Cost-Effectiveness (SpRUCE checklist) were used to assess the reporting and quality of health utilities in glaucoma CUA. Findings 43 CUAs were included, with 11 unique sources of health utilities. A wide range of health utilities for the same Hodapp-Parrish-Anderson glaucoma health states were reported; ocular hypertension (0.84–0.95), mild (0.68–0.94), moderate (0.57–0.92), advanced (0.58–0.88), severe/blind (0.46–0.76), and bilateral blindness (0.26–0.5). Most studies reported the basis for using health utilities (34, 79%) and any assumptions or adjustments applied to the health utilities (22, 51%). Few studies reported a framework for assessing the relevance of health utilities to a decision context (8, 19%). Even fewer (3, 7%) applied a systematic search strategy to identify health utilities and used a structured assessment of quality for inclusion. Overall, reporting has not improved over time. Conclusions and relevance This review describes that few CUAs describe important rationale for using health state utility values. Including additional details on the search, appraisal, selection, and inclusion process of health utility values improves transparency, generalizability and supports the assessment of the validity of study conclusions. Future investigations should aim to use health utilities on the same scale of measurement across health states and consider the source and relevance to the decision context/purpose of conducting that cost-utility study.
- 27 February 2024
- Published by Springer Nature in The European Journal of Health Economics
- Vol. 25 (8) , 1417-1435
Systematic methodological review of health state values in glaucoma cost-utility analyses
Importance Describing the characteristics and sources of health state utility values and reporting practice in the literature of cost-utility analyses facilitates an understanding of the level of the transparency, validity, and generalizability of cost-utility analyses. Improving the quality of reporting will support investigators in describing the incremental value of emerging glaucoma interventions. Objective To describe the state of practice among published glaucoma cost-utility analysis studies, focusing on valuation of health and the quality of reporting. Evidence review We searched several databases including Medline, CINHAL, Embase, Web of Science, Scopus, Biosis previews, the Health Economic Evaluations Database, and the NHS Economic Evaluation Database (NHS EED). We included full-text, English, published cost-utility analyses of glaucoma interventions with quality-adjusted life years (QALYs) as the primary outcome measure to calculate incremental cost-utility ratios. Excluded studies were non-English language, reviews, editorials, protocols, or other types of economic studies (cost–benefit, cost-minimization, cost-effectiveness). Study characteristics, operational definitions of glaucoma health states and health state utilities were extracted. The original source of the health utility was reviewed to determine the scale of measurement and the source of preference weighting. Items from the Systematic Review of Utilities for Cost-Effectiveness (SpRUCE checklist) were used to assess the reporting and quality of health utilities in glaucoma CUA. Findings 43 CUAs were included, with 11 unique sources of health utilities. A wide range of health utilities for the same Hodapp-Parrish-Anderson glaucoma health states were reported; ocular hypertension (0.84–0.95), mild (0.68–0.94), moderate (0.57–0.92), advanced (0.58–0.88), severe/blind (0.46–0.76), and bilateral blindness (0.26–0.5). Most studies reported the basis for using health utilities (34, 79%) and any assumptions or adjustments applied to the health utilities (22, 51%). Few studies reported a framework for assessing the relevance of health utilities to a decision context (8, 19%). Even fewer (3, 7%) applied a systematic search strategy to identify health utilities and used a structured assessment of quality for inclusion. Overall, reporting has not improved over time. Conclusions and relevance This review describes that few CUAs describe important rationale for using health state utility values. Including additional details on the search, appraisal, selection, and inclusion process of health utility values improves transparency, generalizability and supports the assessment of the validity of study conclusions. Future investigations should aim to use health utilities on the same scale of measurement across health states and consider the source and relevance to the decision context/purpose of conducting that cost-utility study.
- Open Access
Prolonged COVID-19 symptom duration in people with systemic autoimmune rheumatic diseases: results from the COVID-19 Global Rheumatology Alliance Vaccine Survey
Objective We investigated prolonged COVID-19 symptom duration, defined as lasting 28 days or longer, among people with systemic autoimmune rheumatic diseases (SARDs). Methods We analysed data from the COVID-19 Global Rheumatology Alliance Vaccine Survey (2 April 2021–15 October 2021) to identify people with SARDs reporting test-confirmed COVID-19. Participants reported COVID-19 severity and symptom duration, sociodemographics and clinical characteristics. We reported the proportion experiencing prolonged symptom duration and investigated associations with baseline characteristics using logistic regression. Results We identified 441 respondents with SARDs and COVID-19 (mean age 48.2 years, 83.7% female, 39.5% rheumatoid arthritis). The median COVID-19 symptom duration was 15 days (IQR 7, 25). Overall, 107 (24.2%) respondents had prolonged symptom duration (≥28 days); 42/429 (9.8%) reported symptoms lasting ≥90 days. Factors associated with higher odds of prolonged symptom duration included: hospitalisation for COVID-19 vs not hospitalised and mild acute symptoms (age-adjusted OR (aOR) 6.49, 95% CI 3.03 to 14.1), comorbidity count (aOR 1.11 per comorbidity, 95% CI 1.02 to 1.21) and osteoarthritis (aOR 2.11, 95% CI 1.01 to 4.27). COVID-19 onset in 2021 vs June 2020 or earlier was associated with lower odds of prolonged symptom duration (aOR 0.42, 95% CI 0.21 to 0.81). Conclusion Most people with SARDs had complete symptom resolution by day 15 after COVID-19 onset. However, about 1 in 4 experienced COVID-19 symptom duration 28 days or longer; 1 in 10 experienced symptoms 90 days or longer. Future studies are needed to investigate the possible relationships between immunomodulating medications, SARD type/flare, vaccine doses and novel viral variants with prolonged COVID-19 symptoms and other postacute sequelae of COVID-19 among people with SARDs.
Open AccessProlonged COVID-19 symptom duration in people with systemic autoimmune rheumatic diseases: results from the COVID-19 Global Rheumatology Alliance Vaccine Survey
Objective We investigated prolonged COVID-19 symptom duration, defined as lasting 28 days or longer, among people with systemic autoimmune rheumatic diseases (SARDs). Methods We analysed data from the COVID-19 Global Rheumatology Alliance Vaccine Survey (2 April 2021–15 October 2021) to identify people with SARDs reporting test-confirmed COVID-19. Participants reported COVID-19 severity and symptom duration, sociodemographics and clinical characteristics. We reported the proportion experiencing prolonged symptom duration and investigated associations with baseline characteristics using logistic regression. Results We identified 441 respondents with SARDs and COVID-19 (mean age 48.2 years, 83.7% female, 39.5% rheumatoid arthritis). The median COVID-19 symptom duration was 15 days (IQR 7, 25). Overall, 107 (24.2%) respondents had prolonged symptom duration (≥28 days); 42/429 (9.8%) reported symptoms lasting ≥90 days. Factors associated with higher odds of prolonged symptom duration included: hospitalisation for COVID-19 vs not hospitalised and mild acute symptoms (age-adjusted OR (aOR) 6.49, 95% CI 3.03 to 14.1), comorbidity count (aOR 1.11 per comorbidity, 95% CI 1.02 to 1.21) and osteoarthritis (aOR 2.11, 95% CI 1.01 to 4.27). COVID-19 onset in 2021 vs June 2020 or earlier was associated with lower odds of prolonged symptom duration (aOR 0.42, 95% CI 0.21 to 0.81). Conclusion Most people with SARDs had complete symptom resolution by day 15 after COVID-19 onset. However, about 1 in 4 experienced COVID-19 symptom duration 28 days or longer; 1 in 10 experienced symptoms 90 days or longer. Future studies are needed to investigate the possible relationships between immunomodulating medications, SARD type/flare, vaccine doses and novel viral variants with prolonged COVID-19 symptoms and other postacute sequelae of COVID-19 among people with SARDs.
- 16 June 2022
- Published by Elsevier in Gastroenterology
- Vol. 162 (7)
Mo1652: CAUSES OF FALSE POSITIVE DETECTIONS BY CADE DURING REAL-TIME COLONOSCOPY AND ITS CLINICAL IMPACT: A SYSTEMATIC REVIEW AND META-ANALYSIS
- 16 June 2022
- Published by Elsevier in Gastroenterology
- Vol. 162 (7)
Mo1652: CAUSES OF FALSE POSITIVE DETECTIONS BY CADE DURING REAL-TIME COLONOSCOPY AND ITS CLINICAL IMPACT: A SYSTEMATIC REVIEW AND META-ANALYSIS
- Open Access
- 23 April 2022
- Published by Oxford University Press (OUP) in Rheumatology
- Vol. 61 (SI2) , SI143-SI150
Baseline factors associated with self-reported disease flares following COVID-19 vaccination among adults with systemic rheumatic disease: results from the COVID-19 global rheumatology alliance vaccine survey
To examine the frequency of, and risk factors for, disease flare following COVID-19 vaccination in patients with systemic rheumatic disease (SRD). An international study was conducted from 2 April to 16 August 2021, using an online survey of 5619 adults with SRD for adverse events following COVID-19 vaccination, including flares of disease requiring a change in treatment. We examined risk factors identified a priori based on published associations with SRD activity and SARS-CoV-2 severity, including demographics, SRD type, comorbidities, vaccine type, cessation of immunosuppressive medications around vaccination and history of reactions to non-COVID-19 vaccines, using multivariable logistic regression. Flares requiring a change in treatment following COVID-19 vaccination were reported by 4.9% of patients. Compared with rheumatoid arthritis, certain SRD, including systemic lupus erythematosus (OR 1.51, 95% CI 1.03, 2.20), psoriatic arthritis (OR 1.95, 95% CI 1.20, 3.18) and polymyalgia rheumatica (OR 1.94, 95% CI 1.08, 2.48) were associated with higher odds of flare, while idiopathic inflammatory myopathies were associated with lower odds for flare (OR 0.54, 95% CI 0.31–0.96). The Oxford-AstraZeneca vaccine was associated with higher odds of flare relative to the Pfizer-BioNTech vaccine (OR 1.44, 95% CI 1.07, 1.95), as were a prior reaction to a non-COVID-19 vaccine (OR 2.50, 95% CI 1.76, 3.54) and female sex (OR 2.71, 95% CI 1.55, 4.72). SRD flares requiring changes in treatment following COVID-19 vaccination were uncommon in this large international study. Several potential risk factors, as well as differences by disease type, warrant further examination in prospective cohorts.
- 23 April 2022
- Published by Oxford University Press (OUP) in Rheumatology
- Vol. 61 (SI2) , SI143-SI150
Open AccessBaseline factors associated with self-reported disease flares following COVID-19 vaccination among adults with systemic rheumatic disease: results from the COVID-19 global rheumatology alliance vaccine survey
To examine the frequency of, and risk factors for, disease flare following COVID-19 vaccination in patients with systemic rheumatic disease (SRD). An international study was conducted from 2 April to 16 August 2021, using an online survey of 5619 adults with SRD for adverse events following COVID-19 vaccination, including flares of disease requiring a change in treatment. We examined risk factors identified a priori based on published associations with SRD activity and SARS-CoV-2 severity, including demographics, SRD type, comorbidities, vaccine type, cessation of immunosuppressive medications around vaccination and history of reactions to non-COVID-19 vaccines, using multivariable logistic regression. Flares requiring a change in treatment following COVID-19 vaccination were reported by 4.9% of patients. Compared with rheumatoid arthritis, certain SRD, including systemic lupus erythematosus (OR 1.51, 95% CI 1.03, 2.20), psoriatic arthritis (OR 1.95, 95% CI 1.20, 3.18) and polymyalgia rheumatica (OR 1.94, 95% CI 1.08, 2.48) were associated with higher odds of flare, while idiopathic inflammatory myopathies were associated with lower odds for flare (OR 0.54, 95% CI 0.31–0.96). The Oxford-AstraZeneca vaccine was associated with higher odds of flare relative to the Pfizer-BioNTech vaccine (OR 1.44, 95% CI 1.07, 1.95), as were a prior reaction to a non-COVID-19 vaccine (OR 2.50, 95% CI 1.76, 3.54) and female sex (OR 2.71, 95% CI 1.55, 4.72). SRD flares requiring changes in treatment following COVID-19 vaccination were uncommon in this large international study. Several potential risk factors, as well as differences by disease type, warrant further examination in prospective cohorts.
- Open Access
Responsiveness of the HUG-5 in an outpatient clinic: a 12-month randomised feasibility study protocol
Introduction Glaucoma is a progressive, chronic condition that can have a significant impact on a patient’s health-related quality of life (HRQoL). Validated, disease-specific HRQoL tools such as the Health Utility for Glaucoma (HUG-5) tool and the Glaucoma Quality of Life Questionnaire (GlauQoL-17) can be used to monitor a patient’s quality of life. However, the utility of these tools in outpatient clinic practice is not well established. The primary objective of this study is to characterise the feasibility of administering periodic HRQoL questionnaires in glaucoma using a semi-automated workflow. Methods and analysis This study will be a single-centre, unblinded, randomised, parallel-group study with an exploratory data analysis framework. We aim to determine the feasibility of administering the HUG-5 in an outpatient clinic using a semi-automated workflow and determine patient engagement through email and telephone contact methods. We will also be investigating the association of the HUG-5 and GlauQoL-17 with patient visual field testing and visual acuity. Mean differences between groups will be tested with analysis of variance to determine if the frequency of calls affects burden, satisfaction and perceived value of information. Ethics and dissemination This study has been approved by the Hamilton Integrated Research Ethics board (ID: 13046) and will be conducted within Canadian Tri-Council Statement policy. Personal information of the study’s participants will be anonymised with identification codes and data will be kept on an encrypted server. Results of this study will be disseminated through peer-reviewed journals, conferences and internal meetings.
Open AccessResponsiveness of the HUG-5 in an outpatient clinic: a 12-month randomised feasibility study protocol
Introduction Glaucoma is a progressive, chronic condition that can have a significant impact on a patient’s health-related quality of life (HRQoL). Validated, disease-specific HRQoL tools such as the Health Utility for Glaucoma (HUG-5) tool and the Glaucoma Quality of Life Questionnaire (GlauQoL-17) can be used to monitor a patient’s quality of life. However, the utility of these tools in outpatient clinic practice is not well established. The primary objective of this study is to characterise the feasibility of administering periodic HRQoL questionnaires in glaucoma using a semi-automated workflow. Methods and analysis This study will be a single-centre, unblinded, randomised, parallel-group study with an exploratory data analysis framework. We aim to determine the feasibility of administering the HUG-5 in an outpatient clinic using a semi-automated workflow and determine patient engagement through email and telephone contact methods. We will also be investigating the association of the HUG-5 and GlauQoL-17 with patient visual field testing and visual acuity. Mean differences between groups will be tested with analysis of variance to determine if the frequency of calls affects burden, satisfaction and perceived value of information. Ethics and dissemination This study has been approved by the Hamilton Integrated Research Ethics board (ID: 13046) and will be conducted within Canadian Tri-Council Statement policy. Personal information of the study’s participants will be anonymised with identification codes and data will be kept on an encrypted server. Results of this study will be disseminated through peer-reviewed journals, conferences and internal meetings.
- Top Cited PapersOpen Access
Early experience of COVID-19 vaccination in adults with systemic rheumatic diseases: results from the COVID-19 Global Rheumatology Alliance Vaccine Survey
Background We describe the early experiences of adults with systemic rheumatic disease who received the COVID-19 vaccine. Methods From 2 April to 30 April 2021, we conducted an online, international survey of adults with systemic rheumatic disease who received COVID-19 vaccination. We collected patient-reported data on clinician communication, beliefs and intent about discontinuing disease-modifying antirheumatic drugs (DMARDs) around the time of vaccination, and patient-reported adverse events after vaccination. Results We analysed 2860 adults with systemic rheumatic diseases who received COVID-19 vaccination (mean age 55.3 years, 86.7% female, 86.3% white). Types of COVID-19 vaccines were Pfizer-BioNTech (53.2%), Oxford/AstraZeneca (22.6%), Moderna (21.3%), Janssen/Johnson & Johnson (1.7%) and others (1.2%). The most common rheumatic disease was rheumatoid arthritis (42.3%), and 81.2% of respondents were on a DMARD. The majority (81.9%) reported communicating with clinicians about vaccination. Most (66.9%) were willing to temporarily discontinue DMARDs to improve vaccine efficacy, although many (44.3%) were concerned about rheumatic disease flares. After vaccination, the most reported patient-reported adverse events were fatigue/somnolence (33.4%), headache (27.7%), muscle/joint pains (22.8%) and fever/chills (19.9%). Rheumatic disease flares that required medication changes occurred in 4.6%. Conclusion Among adults with systemic rheumatic disease who received COVID-19 vaccination, patient-reported adverse events were typical of those reported in the general population. Most patients were willing to temporarily discontinue DMARDs to improve vaccine efficacy. The relatively low frequency of rheumatic disease flare requiring medications was reassuring.
Top Cited PapersOpen AccessEarly experience of COVID-19 vaccination in adults with systemic rheumatic diseases: results from the COVID-19 Global Rheumatology Alliance Vaccine Survey
Background We describe the early experiences of adults with systemic rheumatic disease who received the COVID-19 vaccine. Methods From 2 April to 30 April 2021, we conducted an online, international survey of adults with systemic rheumatic disease who received COVID-19 vaccination. We collected patient-reported data on clinician communication, beliefs and intent about discontinuing disease-modifying antirheumatic drugs (DMARDs) around the time of vaccination, and patient-reported adverse events after vaccination. Results We analysed 2860 adults with systemic rheumatic diseases who received COVID-19 vaccination (mean age 55.3 years, 86.7% female, 86.3% white). Types of COVID-19 vaccines were Pfizer-BioNTech (53.2%), Oxford/AstraZeneca (22.6%), Moderna (21.3%), Janssen/Johnson & Johnson (1.7%) and others (1.2%). The most common rheumatic disease was rheumatoid arthritis (42.3%), and 81.2% of respondents were on a DMARD. The majority (81.9%) reported communicating with clinicians about vaccination. Most (66.9%) were willing to temporarily discontinue DMARDs to improve vaccine efficacy, although many (44.3%) were concerned about rheumatic disease flares. After vaccination, the most reported patient-reported adverse events were fatigue/somnolence (33.4%), headache (27.7%), muscle/joint pains (22.8%) and fever/chills (19.9%). Rheumatic disease flares that required medication changes occurred in 4.6%. Conclusion Among adults with systemic rheumatic disease who received COVID-19 vaccination, patient-reported adverse events were typical of those reported in the general population. Most patients were willing to temporarily discontinue DMARDs to improve vaccine efficacy. The relatively low frequency of rheumatic disease flare requiring medications was reassuring.
- 16 July 2020
- Published by Wiley in Journal of Cardiovascular Electrophysiology
- Vol. 31 (10) , 2712-2719
Cardiac implantable electronic device placement following alcohol septal ablation for hypertrophic cardiomyopathy in the United States
Background: Cardiac implantable electronic devices (CIED) are sometimes required after alcohol septal ablation (ASA) for hypertrophic cardiomyopathy (HCM). The primary objectives of this study were to characterize the incidence, timing, and predictors of CIED placement after ASA for HCM.Methods: Patients were identified from the 2010–2015 Nationwide Readmissions Databases. Incidence, timing and independent predictors of CIED placement, as well as 30‐day readmission rates were examined.Results: There were 1296 patients (national estimate = 2864) with HCM who underwent ASA. CIED were implanted in 322 (25% overall; 14% permanent pacemaker, 11% implantable cardioverter defibrillator) during the index hospitalization. Of these, 21%, 23%, 21%, and 18% occurred on postprocedure day 0, 1, 2, and 3, respectively. Only 17 (1.3%) patients underwent CIED implantation between discharge and 30‐day follow up. Independent predictors of index hospitalization CIED implantation included older age, diabetes, heart failure, nonelective index hospital admission and hospitalization at a privately owned hospital. Nonelective 30‐day readmission rates among those who did and did not undergo CIED placement during their index hospitalization, were 6.8% and 7.9%, respectively (p = .53); median time to readmission was also similar between groups.Conclusions: One in four HCM patients undergoing ASA underwent CIED implantation during their index hospitalization; nearly 2/3rd during the first 48 h postprocedure. Private hospital ownership independently predicted CIED placement. More data are needed to better understand the unexpectedly high rates of CIED placement, earlier than anticipated timing of implantation and differential rates by hospital ownership.
- 16 July 2020
- Published by Wiley in Journal of Cardiovascular Electrophysiology
- Vol. 31 (10) , 2712-2719
Cardiac implantable electronic device placement following alcohol septal ablation for hypertrophic cardiomyopathy in the United States
Background: Cardiac implantable electronic devices (CIED) are sometimes required after alcohol septal ablation (ASA) for hypertrophic cardiomyopathy (HCM). The primary objectives of this study were to characterize the incidence, timing, and predictors of CIED placement after ASA for HCM.Methods: Patients were identified from the 2010–2015 Nationwide Readmissions Databases. Incidence, timing and independent predictors of CIED placement, as well as 30‐day readmission rates were examined.Results: There were 1296 patients (national estimate = 2864) with HCM who underwent ASA. CIED were implanted in 322 (25% overall; 14% permanent pacemaker, 11% implantable cardioverter defibrillator) during the index hospitalization. Of these, 21%, 23%, 21%, and 18% occurred on postprocedure day 0, 1, 2, and 3, respectively. Only 17 (1.3%) patients underwent CIED implantation between discharge and 30‐day follow up. Independent predictors of index hospitalization CIED implantation included older age, diabetes, heart failure, nonelective index hospital admission and hospitalization at a privately owned hospital. Nonelective 30‐day readmission rates among those who did and did not undergo CIED placement during their index hospitalization, were 6.8% and 7.9%, respectively (p = .53); median time to readmission was also similar between groups.Conclusions: One in four HCM patients undergoing ASA underwent CIED implantation during their index hospitalization; nearly 2/3rd during the first 48 h postprocedure. Private hospital ownership independently predicted CIED placement. More data are needed to better understand the unexpectedly high rates of CIED placement, earlier than anticipated timing of implantation and differential rates by hospital ownership.
- Open Access
- 27 February 2020
- Published by Springer Nature in The European Journal of Health Economics
- Vol. 21 (5) , 717-728
Is the whole larger than the sum of its parts? Impact of missing data imputation in economic evaluation conducted alongside randomized controlled trials
Outcomes in economic evaluations, such as health utilities and costs, are products of multiple variables, often requiring complete item responses to questionnaires. Therefore, missing data are very common in cost-effectiveness analyses. Multiple imputations (MI) are predominately recommended and could be made either for individual items or at the aggregate level. We, therefore, aimed to assess the precision of both MI approaches (the item imputation vs. aggregate imputation) on the cost-effectiveness results. The original data set came from a cluster-randomized, controlled trial and was used to describe the missing data pattern and compare the differences in the cost-effectiveness results between the two imputation approaches. A simulation study with different missing data scenarios generated based on a complete data set was used to assess the precision of both imputation approaches. For health utility and cost, patients more often had a partial (9% vs. 23%, respectively) rather than complete missing (4% vs. 0%). The imputation approaches differed in the cost-effectiveness results (the item imputation: − 61,079€/QALY vs. the aggregate imputation: 15,399€/QALY). Within the simulation study mean relative bias (< 5% vs. < 10%) and range of bias (< 38% vs. < 83%) to the true incremental cost and incremental QALYs were lower for the item imputation compared to the aggregate imputation. Even when 40% of data were missing, relative bias to true cost-effectiveness curves was less than 16% using the item imputation, but up to 39% for the aggregate imputation. Thus, the imputation strategies could have a significant impact on the cost-effectiveness conclusions when more than 20% of data are missing. The item imputation approach has better precision than the imputation at the aggregate level.
- 27 February 2020
- Published by Springer Nature in The European Journal of Health Economics
- Vol. 21 (5) , 717-728
Open AccessIs the whole larger than the sum of its parts? Impact of missing data imputation in economic evaluation conducted alongside randomized controlled trials
Outcomes in economic evaluations, such as health utilities and costs, are products of multiple variables, often requiring complete item responses to questionnaires. Therefore, missing data are very common in cost-effectiveness analyses. Multiple imputations (MI) are predominately recommended and could be made either for individual items or at the aggregate level. We, therefore, aimed to assess the precision of both MI approaches (the item imputation vs. aggregate imputation) on the cost-effectiveness results. The original data set came from a cluster-randomized, controlled trial and was used to describe the missing data pattern and compare the differences in the cost-effectiveness results between the two imputation approaches. A simulation study with different missing data scenarios generated based on a complete data set was used to assess the precision of both imputation approaches. For health utility and cost, patients more often had a partial (9% vs. 23%, respectively) rather than complete missing (4% vs. 0%). The imputation approaches differed in the cost-effectiveness results (the item imputation: − 61,079€/QALY vs. the aggregate imputation: 15,399€/QALY). Within the simulation study mean relative bias (< 5% vs. < 10%) and range of bias (< 38% vs. < 83%) to the true incremental cost and incremental QALYs were lower for the item imputation compared to the aggregate imputation. Even when 40% of data were missing, relative bias to true cost-effectiveness curves was less than 16% using the item imputation, but up to 39% for the aggregate imputation. Thus, the imputation strategies could have a significant impact on the cost-effectiveness conclusions when more than 20% of data are missing. The item imputation approach has better precision than the imputation at the aggregate level.
- 1 July 2019
- Published by Wolters Kluwer Health in Journal of Glaucoma
- Vol. 28 (7) , 593-600
Disease-specific Preference-based Measure of Glaucoma Health States: HUG-5 Psychometric Validation
Glaucoma is the second leading cause of irreversible blindness in the world, with 60 million people worldwide estimated to suffer from the condition. Health utility is an important outcome measure of treatment effect in economic outcomes. The available tools for the measurement of health utility in glaucoma patients do not support a preference-based algorithm required to estimate health utility. To resolve this gap in the literature, the HUG-5 (Health Utility for Glaucoma—5 dimensions) was developed. The objective of the present study was to validate the HUG-5 with accepted measures of health state and vision-specific quality of life. The HUG-5 measures patient self-reported levels of visual discomfort, mobility, daily life activities, emotion, and social activities, as affected by the progression and management of glaucoma. To evaluate the psychometric properties, the HUG-5 was assessed for construct validity between similar and dissimilar dimensions of the National Eye Institute Visual Function Questionnaire-25 (NEI-VFQ-25) and the EuroQol’s 5 Dimensions. The HUG-5 was evaluated for test-retest reliability after a 2-week period. The HUG-5 composite distributions of mild/moderate and advanced glaucoma patients were tested for differences to measure sensitivity. A total of 124 patients with glaucoma were administered the NEI-VFQ-25, the HUG-5, and the EuroQol’s 5 Dimensions. The HUG-5 demonstrated construct validity, with convergent and discriminant support for visual discomfort, mobility, daily life activities, emotional distress, and social activities. The HUG-5 concurrently measured health-related quality of life associated with best-eye visual field loss (r=0.63, P<0.001). The HUG-5 measured health state consistently with test-retest reliability (intraclass correlation=0.91, P<0.001). The HUG-5 was established to be sensitive in detecting differences between patients with mild/moderate glaucoma and those with advanced glaucoma with a rank-sum test with continuity correction (W=693.5, P<0.001). This study demonstrates promising results for the HUG-5’s response range and relationship with the NEI-VFQ-25 and best-eye visual field loss, highlighting the value of disease-specific preference-based scoring systems in measuring health state changes in glaucoma patients.
- 1 July 2019
- Published by Wolters Kluwer Health in Journal of Glaucoma
- Vol. 28 (7) , 593-600
Disease-specific Preference-based Measure of Glaucoma Health States: HUG-5 Psychometric Validation
Glaucoma is the second leading cause of irreversible blindness in the world, with 60 million people worldwide estimated to suffer from the condition. Health utility is an important outcome measure of treatment effect in economic outcomes. The available tools for the measurement of health utility in glaucoma patients do not support a preference-based algorithm required to estimate health utility. To resolve this gap in the literature, the HUG-5 (Health Utility for Glaucoma—5 dimensions) was developed. The objective of the present study was to validate the HUG-5 with accepted measures of health state and vision-specific quality of life. The HUG-5 measures patient self-reported levels of visual discomfort, mobility, daily life activities, emotion, and social activities, as affected by the progression and management of glaucoma. To evaluate the psychometric properties, the HUG-5 was assessed for construct validity between similar and dissimilar dimensions of the National Eye Institute Visual Function Questionnaire-25 (NEI-VFQ-25) and the EuroQol’s 5 Dimensions. The HUG-5 was evaluated for test-retest reliability after a 2-week period. The HUG-5 composite distributions of mild/moderate and advanced glaucoma patients were tested for differences to measure sensitivity. A total of 124 patients with glaucoma were administered the NEI-VFQ-25, the HUG-5, and the EuroQol’s 5 Dimensions. The HUG-5 demonstrated construct validity, with convergent and discriminant support for visual discomfort, mobility, daily life activities, emotional distress, and social activities. The HUG-5 concurrently measured health-related quality of life associated with best-eye visual field loss (r=0.63, P<0.001). The HUG-5 measured health state consistently with test-retest reliability (intraclass correlation=0.91, P<0.001). The HUG-5 was established to be sensitive in detecting differences between patients with mild/moderate glaucoma and those with advanced glaucoma with a rank-sum test with continuity correction (W=693.5, P<0.001). This study demonstrates promising results for the HUG-5’s response range and relationship with the NEI-VFQ-25 and best-eye visual field loss, highlighting the value of disease-specific preference-based scoring systems in measuring health state changes in glaucoma patients.