The research team, working in collaboration with investigators from MD Anderson Cancer Center, Yale University, and Columbia University, tracked more than 12,000 prescriptions for oral blood cancer drugs in 2022 among patients insured through Medicare or commercial health plans. Rather than simply measuring whether patients eventually obtained their medications, the investigators followed each prescription along a complete pathway: from the initial submission at the pharmacy, through the insurer’s approval or rejection decision, to whether the patient ultimately filled the prescription within 90 days. This longitudinal design allowed the researchers to quantify not just a single barrier but a cascade of them, revealing two distinct points at which patients fall away from treatment.

The first drop-off occurs at the pharmacy counter before treatment even begins. At the initial submission, 64.9 percent of prescriptions for Medicare patients and a striking 84 percent of prescriptions for commercially insured patients were rejected by insurers. The most common reasons were prior authorization requirements, in which a prescriber must submit additional documentation to justify the drug, and outright non-coverage, in which the medication simply was not included in the patient’s plan formulary. Among Medicare beneficiaries, 32.1 percent of prescriptions were initially rejected because prior authorization was required, while among commercially insured patients, 27.8 percent were rejected for prior authorization and 17.1 percent because the drug was not covered at all.

Time and persistence resolved many of these rejections. Within 90 days, the share of prescriptions still rejected had fallen dramatically: insurer approval rates climbed to 85 percent for Medicare patients and 62.9 percent for commercially insured patients. Prior authorization rejections in particular proved largely surmountable, falling from 32.1 percent to just 4.6 percent among Medicare patients, and from 27.8 percent to 6.9 percent among commercially insured patients. Non-coverage rejections also declined, dropping to 8.0 percent in the commercial group. The researchers caution that they could not determine whether any individual rejection was clinically appropriate, but the pattern is telling: when insurers that initially refused a drug ultimately approve the very same therapy, the initial refusal appears to have imposed administrative work and treatment delay without changing the clinical outcome.

For patients with aggressive blood cancers, such delays are not trivial. Targeted oral therapies for chronic myeloid leukemia, for example, work best when started promptly, and clinicians generally regard rapid treatment initiation as essential for diseases that can progress over weeks or months. Every day spent navigating paperwork is a day the underlying malignancy continues to advance. The study’s authors argue that the sheer volume of initial rejections, most of which are later reversed, suggests that prior authorization may function less as a clinical safeguard and more as a friction-generating gatekeeper, one whose costs are borne by patients, oncology practices, and pharmacy staff alike.

The second drop-off is financial, and it occurs after the insurer has said yes. Even with approval in hand, only 54.5 percent of Medicare patients and 45.5 percent of commercially insured patients had a prescription that was both approved and filled within 90 days. The analysis found a steep, nearly linear relationship between out-of-pocket cost and the likelihood that a patient would actually fill the prescription. Among Medicare beneficiaries, fill rates were 84.9 percent when out-of-pocket costs were 15 dollars or less, and 76.7 percent when costs ranged from just over 15 dollars to 175 dollars. But when costs climbed into the 175-to-500-dollar range, fill rates collapsed to 38 percent, and they fell further to 29.2 percent when costs ranged from just over 500 dollars to 2,000 dollars.

Commercially insured patients showed the same pattern, with even steeper penalties at the high end. Fill rates were 80 percent when out-of-pocket costs were 15 dollars or less and 71.3 percent in the 15-to-175-dollar band, but only 30.9 percent when costs ranged from just over 500 to 2,000 dollars, and a mere 21.6 percent when costs exceeded 2,000 dollars. In other words, roughly four out of five patients facing modest copays filled their prescriptions, while fewer than one in four did so when confronted with four-figure cost-sharing. Because specialty oral anticancer drugs are typically covered under pharmacy benefits rather than medical benefits, patients encounter these cost-sharing requirements directly and often unexpectedly, sometimes at the very moment they are processing a frightening new diagnosis.

Senior author Jalpa A. Doshi, the Leon Hess Professor of Internal Medicine and a senior fellow at the Leonard Davis Institute of Health Economics at the University of Pennsylvania, framed the findings as a map of where the prescription pathway fails. She noted that these blood cancer medications can be life-prolonging, but that insurance coverage requirements and high out-of-pocket costs can delay or limit access, and that the national benchmark her team constructed helps identify exactly where patients encounter barriers along the way. Doshi also emphasized that getting an insurer to approve a cancer drug is only the first hurdle, describing a second drop-off at the pharmacy counter when patients face high out-of-pocket costs. For Medicare beneficiaries, she pointed to the new annual out-of-pocket cost cap and the Medicare Prescription Payment Plan, which allows patients to spread prescription costs across the year, as potentially meaningful relief, though she stressed that awareness and enrollment will be critical if patients are to benefit.

The comparison between insurance types carries its own policy implications. Commercially insured patients fared worse on nearly every measure: higher initial rejection rates, lower final approval rates, and lower combined approval-and-fill rates. The authors suggest that reforms to prior authorization processes, drug coverage rules, and cost-sharing structures may be needed to ensure that a patient’s access to oral cancer drugs is not determined by the arbitrary design of the insurance plan they happen to hold. Such reforms could include standardizing and expediting prior authorization for oncology medications, requiring formularies to cover at least one drug within each therapeutic class, and capping specialty drug cost-sharing for commercially insured patients in ways that mirror the protections now phasing in for Medicare.

For a field that has celebrated the scientific triumph of targeted oral oncology, the study is a sobering reminder that a molecule’s efficacy is meaningless if it never reaches the bloodstream. The technical achievements are real: kinase inhibitors, immunomodulators, and proteasome inhibitors taken as daily pills have converted once-fatal diagnoses into manageable chronic conditions for many patients. But the delivery system wrapped around those molecules, with its prior authorization forms, formulary exclusions, and deductible-driven cost-sharing, is failing a substantial fraction of the very patients the drugs were designed to help. The Penn-led analysis turns that failure into measurable, comparable numbers, giving policymakers, insurers, and advocacy organizations a concrete benchmark against which future reforms, from streamlined authorization to expanded payment plans, can be judged. Whether the next generation of blood cancer patients will face the same gauntlet will depend on how quickly those reforms move from the pages of journals into the mechanics of insurance plans.

Subject of Research: Insurance coverage barriers and out-of-pocket costs affecting access to oral blood cancer medications

Article Title: Insurance barriers limit access to blood cancer drugs

Article References: Insurance barriers limit access to blood cancer drugs. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: blood cancer, oral anticancer drugs, prior authorization, insurance coverage, out-of-pocket costs, Medicare, prescription fill rates, leukemia, lymphoma, multiple myeloma, pharmacy benefits, health policy

The research, published in the journal Supportive Care in Cancer, enrolled 203 patient-caregiver dyads in which the patient had a hematologic malignancy. Each member of the pair independently completed the EuroQol 5-Dimensions 5-Levels questionnaire, known as the EQ-5D-5L, one of the most widely used instruments in health-outcomes research. The tool asks respondents to rate themselves on five dimensions of health: mobility, self-care, usual activities, pain or discomfort, and anxiety or depression, each on a five-level severity scale. Responses can be summarized as a single utility score, a number between 0 and 1 that anchors full health at 1 and death at 0, making the measure useful both for bedside assessment and for health-economic calculations such as quality-adjusted life years.

The central question was straightforward: when a caregiver answers these questions on a patient’s behalf, does the answer resemble what the patient would have said? The answer, by the statistical standards of psychometrics, was largely yes. Caregiver proxy ratings of the EQ-5D-5L utility score correlated strongly with patient self-ratings, with a correlation coefficient of 0.679, a value conventionally interpreted as a strong positive association. At the level of individual dimensions, correlations ranged from 0.518 to 0.695, all statistically significant. Agreement, a stricter test than correlation because it penalizes systematic bias, was also solid: the intraclass correlation coefficient for utility scores was 0.730, with a 95 percent confidence interval of 0.658 to 0.788, a range that researchers typically classify as good reliability.

Dimension-by-dimension, the picture held up. Exact agreement between what caregivers reported and what patients reported about themselves ranged from 66.5 percent to 84.2 percent across the five EQ-5D-5L domains. In practical terms, when a caregiver marked a patient as having no problems walking, or severe pain, or difficulty with usual activities, that judgment matched the patient’s own answer roughly two-thirds to more than four-fifths of the time. For a population in which fatigue, cytopenias, and treatment toxicity often make self-report impossible, the researchers conclude that caregivers may serve as reliable proxies for assessing health-related quality of life in patients with hematologic malignancies.

Yet the study also uncovered a systematic wrinkle that clinicians should not ignore. Caregivers who were themselves struggling emotionally tended to paint a darker picture of their relative’s condition than the patients did. Greater anxiety symptoms among caregivers, measured with the seven-item Generalized Anxiety Disorder scale, were significantly associated with overestimation of the patients’ symptom burden, with a correlation coefficient of 0.408. Depressive symptoms, assessed with the nine-item Patient Health Questionnaire, showed a similar but weaker association, at 0.288. Both relationships were highly statistically significant. In other words, an anxious or depressed caregiver watching a loved one through chemotherapy may project some of that distress onto the patient’s own experience, rating symptoms as worse than the patient reports them to be.

This phenomenon, known in the literature as the proxy-rating bias, has been documented across cancer care and beyond. Previous systematic reviews of caregiver responses for patient quality-of-life assessment in adult oncology have found that proxies generally track patient self-reports reasonably well but tend to overestimate physical and emotional symptom burden. Studies in glioma patient-caregiver dyads have similarly shown that the psychosocial functioning of the rater influences rating accuracy, and research in dementia care has revealed substantial discrepancies between self- and proxy-rated quality of life. The new findings extend this evidence into hematologic malignancies, a population that has historically been underrepresented in quality-of-life research compared with solid tumors, partly because the episodic and often acute nature of blood cancers complicates longitudinal patient-reported outcome collection.

The methodological design of the study reflects careful attention to measurement quality. Beyond the EQ-5D-5L, caregivers completed the Family APGAR Index, a brief instrument capturing their perception of family functioning across adaptability, partnership, growth, affection, and resolve. The choice of the five-level version of the EQ-5D rather than the older three-level version matters as well: the five-level format reduces ceiling effects and improves discrimination among patients with mild to moderate impairment, which is essential when the goal is to detect subtle differences between raters. The Chinese EQ-5D-5L value set was used to compute utility scores, anchoring the analysis in a validated preference-based framework. Agreement statistics followed established conventions, with intraclass correlation coefficients interpreted according to widely accepted guidelines and categorical agreement benchmarked against classic standards for observer agreement research.

The clinical implications are twofold. First, the results legitimize a pragmatic workflow: when a patient with a hematologic malignancy cannot complete a quality-of-life assessment, whether because of severe illness, cognitive impairment, or the sheer exhaustion of intensive therapy, a well-informed caregiver’s report can stand in with acceptable fidelity. This matters for symptom monitoring, which randomized trials in oncology have shown can improve quality of life and even survival when patient-reported outcomes are systematically collected during routine treatment. It also matters for health-economic evaluation, where proxy utility scores are often needed to estimate the value of new therapies for blood cancers, a field in which cost-effectiveness analysis has grown rapidly. Second, the findings argue for screening caregivers themselves. Because caregiver anxiety and depression distort proxy ratings, treating the caregiver’s psychological distress is not only an act of compassion but also a way to sharpen the accuracy of the clinical data the care team depends on.

The study was approved by the Xiangya Hospital Ethics Committee with written informed consent obtained in accordance with the Declaration of Helsinki, and the authors report no competing interests. The work was supported by the Hunan Provincial Natural Science Foundation of China and the China Postdoctoral Science Foundation. The research team, led by corresponding author Yajing Xu with first author Wei Qin, notes that data are available from the lead and corresponding authors upon reasonable request. As blood cancers continue to impose a rising global burden, with incidence and disability from hematologic malignancies climbing over the past three decades, the message of this study is quietly empowering: the people who know patients best, and who sit beside them through the hardest days, can be trusted to give voice to what those patients are too sick to say, provided clinicians remember to ask how the storytellers themselves are doing.

Subject of Research: Caregiver proxy assessment of health-related quality of life in patients with hematologic malignancies

Article Title: Caregivers’ proxy assessments of health-related quality of life in patients with hematologic malignancies

Article References: Caregivers’ proxy assessments of health-related quality of life in patients with hematologic malignancies. (n.d.). https://doi.org/10.1007/s00520-026-11206-8

Image Credits: AI Generated

DOI: 10.1007/s00520-026-11206-8

Keywords: hematologic malignancies, caregivers, proxy assessment, health-related quality of life, EQ-5D-5L, blood cancer, patient-reported outcomes, anxiety, depression, family functioning, symptom burden, supportive care