Research Letter
January 2, 2025
Knowledge and Worry Following Review of Standard vs Patient-Centered Pathology Reports
Cathryn J. Lapedis, Sophia R. Kurnot, Sarah E. Bergholtz, et al
JAMA. Published online January 2, 2025. doi:10.1001/jama.2024.25461
The immediate release of test results to patients via patient portals is required.1,2 Unfortunately, pathology reports contain complex medical terminology, are not written for patients, and are often read by patients before discussion with their clinician.3,4 Whether patients can extract relevant diagnostic knowledge from such reports is unclear. To address this challenge, researchers have designed patient-centered pathology report (PCPR) formats, which present the most important clinical data from the pathology report in plain language.5 Pathologists can generate PCPRs as a supplement to their standard report using a template in a few minutes. However, few studies of PCPRs exist, and no previous study has directly compared PCPRs with standard formats in current use.6 This study compared diagnosis knowledge and worry among adults presented with different formats of prostate biopsy reports.
Methods
We recruited an online sample of US adults aged 55 to 84 years from Forthright Access, using iterative rounds of demographically targeted invitations to achieve a desired sample (Supplement 1). We recruited individuals who had a prostate and no history of prostate cancer to mirror the rate of new prostate cancer cases in terms of race and ethnicity. Invitations did not include any survey content information.
Participants read a hypothetical scenario in which they presented to the urologist for trouble urinating, underwent a prostate biopsy procedure, were prepared for the range of possible biopsy results as they would be by a typical urologist, and received notification of their pathology report in the patient portal. Participants were randomized to receive 1 of 3 report formats (a PCPR adapted from the study by Nayak et al,5 a standard university format, or a standard Veterans Affairs [VA] format) and 1 of 2 risk levels (high: Gleason score of 4 + 4 = 8; low: Gleason score of 3 + 3 = 6) (Supplement 2). All formats contained the information needed to answer the knowledge questions. Diagnosis knowledge items included participants’ ability to identify the diagnosis (prostate cancer), the total Gleason score, and the risk level. Information on degree of worry (assessed using a 5-point Likert scale ranging from 1 [not worried] to 5 [extremely worried]), ease of understanding the report (assessed using a 6-point Likert scale ranging from 1 [strongly disagree] to 6 [strongly agree]), and self-reported patient demographics were collected. Because online samples are not fully representative, analyses focused on comparisons across the randomized groups, and all participants who did not complete the primary outcome variable questions were excluded from analysis. Pairwise χ2tests were used to compare the observed rates across study outcomes by assigned report. Statistical analysis was completed in Stata SE 18.5 (StataCorp), with a threshold of significance of P < .05 (2-tailed). This study received exempt status (anonymous survey data) from the University of Michigan institutional review boards.
Results
Participants included 2238 adults (mean age, 65.1 years; 3.2% of participants were Asian, 20.5% were Black, 15.2% were Hispanic, 73.8% were White, and 8.3% identified as Other) who met eligibility criteria (Table 1). Completion rates were 799 of 937 participants (85.3%) for the PCPR group, 706 of 927 (76.2%) for the university group, and 733 of 934 (78.5%) for the VA group (P < .001). While 93% of participants who received the PCPR accurately identified that the report showed prostate cancer, only 39% of those who received the university report and 56% of those who received the VA report did so (P < .001). Similarly, 93% of PCPR recipients accurately classified the report as showing either low or high risk vs 41% of university recipients and 36% of VA recipients (P < .001). Accuracy in reporting total Gleason score also differed by format (84% for the PCPR group vs 48% for the university group and 40% for the VA group; P < .001). Reported worry was significantly associated with risk level for PCPR recipients, with a higher level of worry in the high-risk scenario and a lower level of worry in the low-risk scenario compared with the other 2 conditions (P < .001) (Table 2). Ease of understanding ratings was significantly higher for the PCPR group than the standard groups (P < .001).
Discussion
Most study participants could not extract basic information—including whether they have cancer—from standard prostate cancer pathology reports but were able to understand this diagnostic information from the PCPRs. Also, they discriminated between risk levels (ie, lower levels of perceived worry in the low-risk condition) with PCPRs compared with standard reports. Limitations of this study include the use of a nonprobability online sample, a hypothetical scenario, and inability to control for confounding. Hospital systems should consider including PCPRs with standard pathology reports to improve patient understanding.
