Journal of Health Economics and Outcomes Research最新文献

Background: Conservative, noninvasive musculoskeletal treatment delivered digitally has demonstrated similar or better effectiveness in managing and reducing chronic back pain, compared to in-person care. However, there is limited evidence whether digital care reduces future spinal diagnostic imaging visits.

Objectives: The primary goal was to examine the associations between participating in a digital conservative musculoskeletal care program for back pain and subsequent spinal diagnostic imaging use.

Methods: Using medical claims data from a US commercial health plan database, this retrospective, secondary data analysis compared spinal diagnostic imaging visits among a group of digital program participants who had over 12 weeks of back pain to matched patients who only had usual care to treat their back pain. To mitigate selection bias, a propensity score matching model was developed to match study participants based on demographic, comorbidity, baseline medical care use and cost. The study outcomes were any spinal diagnostic imaging visit and number of spinal diagnostic imaging visits per 1000 participants up to 1 year after participating in the digital program.

Results: The study included 2165 digital participants and 2165 matched comparison group patients. We found that digital participants had fewer spinal diagnostic imaging visits in the year after participating in the digital program compared with comparison group patients (14.2% vs 18.2%, P = .0003). The association between the digital program participation and spinal diagnostic imaging visit is stronger in the group who had imaging in the 12 months before, compared to those who had not (-4.8%, P = .007 vs -3.4%, P = .0163).

Discussion: Consistent with previous studies demonstrating that early conservative management is associated with lower odds of imaging, findings from this study offer an encouraging direction for effective alternatives for managing back pain, improving performance outcomes and reducing premature utilization of healthcare services.

Conclusion: The study provides evidence that participating in a digital musculoskeletal program that delivers conservative care is associated with fewer imaging use, especially among participants who had received imaging previously.

Background: Health-related quality of life measurement in infants and toddlers is increasingly important, but generic preference-weighted instruments lack evidence. This study compared the experimental EuroQol Toddler and Infant Populations (EQ-TIPS) and PedsQL in children 0 to 4 years.

Methods: EQ-TIPS-3L v2.0 and PedsQL response distributions were compared by frequency. Item and dimension/summary score associations were computed using Pearson and intra-class correlation coefficient. Age and severity groups (EQ VAS ≥80) were compared with Mann-Whitney U tests.

Results: Cross-sectional data from 260 children were analyzed: 0 to 24 months (n = 111) and 2 to 4 years (n = 149). Most caregivers were mothers, spending significantly more time (≥10 hours) with younger children χ2 = 18.12, P = .001). The EQ-TIPS-3L had the highest problems with eating (27%-31%) and pain (23%-25%) across age groups, with minimal missing data (≤1%). Younger children most frequently had problems with PedsQL: "tired" (54%), "resting a lot" (52%), "crying or fussing when left alone" (61%) and "difficulty soothing when upset" (51%). Older children's main problems were "hurts or aches" (54%), "afraid or scared" (53%), "sad or blue" (50%), "angry" (64%) and "missing school" (56%-65%). All 3 of the PedsQL school items had missing data for older children (27%-30%). Hypothesized item correlations were reached for 30 of 35 and 11 of 12 items in the younger and older groups, respectively. EQ-TIPS-3L LSS showed moderate to strong correlations with all PedsQL scores except for cognitive (0-24 months) and school functioning (2-4 years). Both measures significantly differentiated by severity groups (EQ VAS ≥80) but not by age group.

Conclusion: Both measures showed similar response distributions despite different time frames and response scales. EQ-TIPS-3L eating and pain reported high problems, with eating strongly associated only with PedsQL physical symptoms. The 2- to 4-year PedsQL version had many missing school functioning items; the 13- to 24-month PedsQL may suit older 2- to 4-year-olds better. Low association between PedsQL cognitive functioning and EQ-TIPS-3L suggests further research is needed on this potentially missing construct.

Background: Insomnia affects up to one-third of US adults and is a significant health challenge with an estimated economic burden of up to $100 billion annually. Cognitive behavioral therapy (CBT) for insomnia (CBT-I) is the recommended first-line treatment, but access is limited due to a shortage of trained therapists. Digital CBT-I offers an effective alternative that may enhance accessibility and reduce higher healthcare costs associated with insomnia.

Objective: To evaluate the US healthcare cost-savings of digital CBT-I compared with standard-of-care control.

Methods: A retrospective difference-in-differences analysis compared 1-year preinitiation and post-initiation healthcare costs for 11 027 individuals receiving SleepioRx (FDA-cleared digital CBT treatment for insomnia disorder) compared with 1:1 exact matched controls with insomnia receiving standard care (n = 10 770). Commercial and Medicare claims were adjusted for comorbidities, index year, and baseline utilization.

Results: Digital CBT-I was associated with statistically significant mean annual total cost savings of $2083(95$ 1508-$2657, P < .001) per person, equating to a 42% reduction in costs with SleepioRx relative to matched controls who received standard of care (medications for insomnia).

Discussion: Digital CBT-I was associated with substantial cost savings for payers. The integration of guideline-concordant treatment through digital delivery into standard care pathways offers a promising strategy to address the clinical and economic challenges of insomnia, supporting more efficient resource allocation.

Conclusions: Findings suggest that implementing digital CBT-I at scale may lead to decreased costs for healthcare payers, relative to the current standard of care, while improving access to effective insomnia treatment.

Background: Chemotherapy is the main treatment for patients with metastatic triple-negative breast cancer (mTNBC) who are ineligible for immunotherapy. TNBC is associated with poorer treatment outcomes than other breast cancer subtypes.

Objective: To evaluate treatment patterns, quantify real-world healthcare costs and assess the burden of clinical events of interest (CEIs) among US patients with mTNBC who did not receive immunotherapy.

Methods: This retrospective study used IQVIA PharMetrics® Plus healthcare claims data. Treatment-based proxies were used to identify patients first diagnosed with mTNBC from March 2017 to September 2023. Treatment regimens, frequency and incidence of CEIs, and all-cause, breast cancer-related, and CEI-related costs per patient per month (PPPM, including drug costs) were described during overall follow-up (any line of therapy [LOT]) and during LOT1 and LOT2.

Results: A total of 2717 patients with mTNBC (99.1% female; mean±SD age, 55.6 ± 10.7 years) were identified. Over the follow-up period (median [Q1,Q3], 11.7 [6.0, 26.5] months), most patients (73.1%) only reached LOT1, and the remaining 26.9% of patients had multiple LOTs. Most patients had chemotherapy in LOT1 (98.1%) and LOT2 (90.6%); 98.5% had chemotherapy across any LOTs. Taxanes were the most common, observed in 74.8% of the overall cohort, followed by anthracyclines (56.4%). Across any LOT, 76.1% of patients had ≥1 CEI, most commonly hematological (49.5%), gastrointestinal (44.3%), infusion-related reactions (31.2%), and fatigue (27.8%). Mean (SD) all-cause total costs PPPM were $14\hspace{0.17em}245($ 12 776) overall (of which 79.6% were BC-related and 34.7% were CEI-related), $17\hspace{0.17em}809($ 18 806) during LOT1, and $19\hspace{0.17em}797($ 24 763) during LOT2.

Discussion: Our study findings confirm previously reported high economic burden of mTNBC, with about 80% related to BC treatment. Most patients experienced CEIs during treatment, and these accounted for one-third of their total healthcare costs.

Conclusions: Our study focused on patients with mTNBC with the greatest unmet need, namely those ineligible for immunotherapy. In patients with mTNBC, most of whom received chemotherapy as standard of care, CEIs presented both a clinical and economic burden, highlighting the need for newer treatments that balance total costs of care with adverse events and clinical benefit.

Background: Japan has a unique drug pricing system that in principle reimburses all regulatory-approved drugs. To ensure sustainability, a health technology assessment (HTA) system was introduced in 2019 to adjust the prices of highly innovative and high-budget-impact drugs based on post-reimbursement cost-effectiveness evaluations.

Objectives: This study aimed to examine the nature and contributing factors of differences between manufacturers' and public (the Center for Outcomes Research and Economic Evaluation for Health [C2H]) cost-effectiveness analyses for 31 products evaluated under the Japanese HTA system by March 2025.

Methods: We conducted descriptive analyses comparing manufacturers' and C2H analyses using publicly available reports. Differences in the assessments of additional benefits, incremental cost-effectiveness ratios (ICERs), and reanalysis items were investigated. We explored issues related to orphan drugs and products granted usefulness premiums for attributes not fully captured by quality-adjusted life-years (QALYs), such as improved convenience and prolonged effect.

Results: Among 74 analysis populations across 31 products, 48.6% showed inconsistencies between the manufacturers and C2H in the assessment of additional benefits, outcome measures, or analysis methods used to support those assessments. Inconsistencies in outcome measures and methods increased after the revision of the Japanese HTA system and its guidelines in April 2022. ICER differences were often linked to differences in quality-of-life (QOL) parameters and baseline assumptions. Products granted usefulness premiums for attributes not fully captured by QALYs showed greater ICER differences between the manufacturers and C2H than those without. Although manufacturers often rely on indirect treatment comparisons when evaluating orphan drugs due to limited data and the lack of comparators in clinical trials, these methods were less frequently accepted by C2H due to their associated uncertainty.

Discussion: The findings highlight differences between the manufacturers' and C2H analyses, including evaluation of QOL, orphan drugs, and attributes not captured by QALYs. Providing clearer guidance, considering other countries' HTA systems, may help improve consistency in assessments.

Conclusions: This study identified key differences and contributing factors under the Japanese HTA system. The findings are expected to inform future refinements of the system and its guidelines, thereby promoting more transparent and predictable evaluations.

Background: Long-term inappropriate use of proton pump inhibitors (PPIs) for the treatment of gastroesophageal reflux disease (GERD) and dyspepsia is a common issue that can lead to unnecessary healthcare costs and potential adverse effects. Alternative treatments, such as episodic alginate therapy, may offer a more cost-effective and clinically appropriate approach.

Objective: To investigate the financial impact of transitioning patients from long-term inappropriate PPI use to episodic alginate treatment for GERD and dyspepsia, from the perspective of the National Health Service (NHS) England.

Methods: A budget impact model was used to compare costs over a 5-year period for adult patients using long-term inappropriate PPIs, with and without alginate treatment. By the fifth year, 20% of patients were assumed to have switched to alginate treatment. In this model, the base case analysis included only drug costs, while a scenario analysis also considered adverse effect costs.

Results: Over the 5-year period, net savings of £11.5 million were observed in drug acquisition costs when 20% of patients (4.8 million) successfully transitioned to alginate treatment. When adverse effect costs were included in the scenario analysis, net savings increased to £16.6 million due to a slight reduction in the number of adverse effects. One-way sensitivity analysis confirmed the robustness of these results.

Conclusions: Transitioning patients from long-term PPI use to episodic alginate treatment is beneficial for patients, potentially reducing adverse effects, and can lead to significant budgetary cost savings, which can be reallocated.

Introduction: Economic evaluations are essential for informed healthcare decision-making but often face challenges due to inconsistent reporting and methodological complexity. Large Language Models (LLMs) offer a scalable alternative for evaluating adherence to such standards. Building on Hileas, a previously developed tool, this study assesses the accuracy of LLM-generated evaluations compared with human reviewers, aiming to quantify reliability, identify limitations, and advance automated, but assistive quality assessment methods in health economic research.

Methods: In all, 110 peer-reviewed economic evaluation papers were evaluated using the CHEERS checklist through structured LLM prompts and scored by 2 human reviewers on a 0-4 ordinal scale. Interrater agreement and LLM performance were measured using Cohen's kappa, sensitivity, specificity, and area under the curve. LLM outputs were compared against human consensus ratings, and usability of the review platform was assessed with the System Usability Scale.

Results: Among 2860 item-level evaluations, 25.3% showed disagreement between human reviewers, with generally low interrater reliability (kappa=-0.07 to 0.43). Compared with human consensus, the LLM achieved 72.3% to 94.7% agreement, with areas under the curve up to 0.96 but variable performance across checklist items. At the paper level, LLM-assigned CHEERS scores (median, 17) were consistently lower than human-reviewed scores (median, 18-21).

Conclusion: This study demonstrated an exploratory proof-of-concept application of LLMs to research quality evaluation. Our results suggests that the LLM was generally able to provide well-reasoned evaluations that closely aligned with human assessments, although with some limitations in fully supporting its judgments.

Background: Antiretroviral therapies (ART) have significantly improved the life expectancy of people with HIV (PLWH). However, chronic immune activation and some ART regimens may increase the prevalence of non-HIV comorbidities, such as cardiovascular, renal, bone, and neuropsychiatric conditions. These comorbidities increase healthcare resource utilization and costs for the Spanish National Health System (NHS), yet data on their economic impact remain scarce. Objective: To estimate the healthcare resource use and costs associated with cardiovascular, renal, bone, and neuropsychiatric comorbidities in PLWH from the Spanish NHS perspective and to simulate the financial impact of a potential prevalence increase due to ART toxicity. Methods: An Excel-based model was used to compare a current scenario using national epidemiological data and an alternative scenario with increased comorbidity prevalence due to ART toxicity. Two cohorts were analyzed: PLWH diagnosed for less than 10 years and those diagnosed for 10 years or more. Epidemiological and healthcare utilization data were collected from the literature and validated by an expert panel. Direct healthcare costs, including hospitalizations, tests, medical visits, and emergency care, were estimated and discounted at a 3% annual discount rate. Results: In 2024, 139 390 PLWH would be living in Spain, with 17 046 having cardiovascular, 7752 renal, 17 700 bone, and 16 207 neuropsychiatric comorbidities, predominantly affecting patients diagnosed for at least 10 years. By 2034, these figures will rise to 33 555, 15 391, 33 950, and 27 388, respectively, with increases observed in both cohorts. Estimated 2024 healthcare costs were €83 million, €48 million, €55 million, and €97 million for cardiovascular, renal, bone, and neuropsychiatric comorbidities, respectively. The alternative scenario with increased comorbidities prevalence projected an additional €900 million to €1400 million. Discussion: The projected increase in the prevalence of cardiovascular, renal, bone, and neuropsychiatric comorbidities among PLWH represents a significant challenge for the Spanish NHS, primarily driven by long-term use of specific ART regimes associated with higher toxicity profiles. Conclusion: Non-HIV comorbidities pose a growing economic challenge. Selecting lower-toxicity ART regimens and preventive strategies will be crucial to mitigating financial impact.

Background: Meningitis, pneumonia, and bacteremia, prevalent community-acquired diseases that can lead to multi-organ failure, are influenced by age, comorbidities, and living conditions. Despite meningitis surveillance in Colombia, information on pneumonia and bacteremia remains limited. This study aims to determine frequency of these diseases among Colombian patients and estimate related healthcare resources.

Objective: To measure the frequency and mortality of meningitis, pneumonia, and bacteremia in all diagnosed adult patients in Colombia from 2015 to 2022.

Methods: This retrospective study analyzed adult (≥18 years) patients, from structured data collection (International Classification of Diseases, Tenth Revision) in a health maintenance organization (HMO). Diagnosis of a first meningitis, pneumonia, or bacteremia episode-unknown cause (bacterial etiology undetermined)-between 2015 and 2022 was reviewed. Index date was defined as when the diagnosis was registered. Frequency was calculated by dividing the number of cases by the number of members in the HMO system over the study period.

Results: Among 112 205 patients, 96.0% had pneumonia, 6.2% bacteremia, and 0.4% meningitis, not mutually exclusive. Inpatient pneumonia incidence, which peaked in 2019 and dropped post-COVID pandemic, was 167 cases per 100 000 person-years in 2022. Incidence of meningitis, pneumonia, and bacteremia was higher in patients over 60 years. Common comorbidities were chronic obstructive pulmonary disease and cardiovascular disease. Bacteremia incidence decreased from 143 cases per 100 000 in 2015 to 69.6 in 2022. Meningitis incidence dropped from 5.3 to 2.2 cases per 100 000 in the COVID period. All-cause mortality rates were 12.0%, 33.5% and 13.8% for pneumonia, bacteremia, and meningitis, respectively.

Discussion: This study is the first to use health electronic databases from an HMO to estimate the burden of these diseases in Colombian patients. Incidence was consistent with COVID-period patterns observed in other studies. Mortality rates were higher with bacteremia. Comorbidities like chronic pulmonary disease, cardiovascular disease, kidney diseases, and dementia were linked with increased incidence and mortality, emphasizing the need for targeted healthcare interventions and vaccination programs.

Conclusion: Incidence and mortality, whether pneumonia (inpatient or outpatient), bacteremia, or meningitis with bacteremia, vary with age and comorbidities, while all-cause mortality was greater for bacteremia than pneumonia or meningitis.