Developing Patient-Reported Outcome Measures of Timely Experience of Diagnosis (PROMOTE-Dx) for Cancer: Survey and Quality Measure Development Report

Short abstract

This study documents the creation, testing, and performance of survey-based patient-reported measures for timeliness of cancer diagnosis, the Patient-Reported Outcome Measures of Timely Experience of Diagnosis (PROMOTE-Dx) for Cancer and the survey instrument used to derive these measures. PROMOTE-Dx can be used by health care organizations and health plans seeking to assess and improve timely diagnosis of cancer.

Abstract

This study documents the creation, testing, and performance of survey-based patient-reported measures for timeliness of cancer diagnosis, the Patient-Reported Outcome Measures of Timely Experience of Diagnosis (PROMOTE-Dx) for Cancer and the survey instrument used to derive these measures. PROMOTE-Dx can be used by health care organizations and health plans seeking to assess and improve timely diagnosis of cancer.

Background

Timely diagnosis of cancer is associated with diagnosis at earlier stages, longer survival, and better quality of life (Neal et al., 2015). The cancer diagnostic process is composed of three distinct intervals (Scott et al., 2013):

1. In the self-appraisal interval, patients may detect symptoms or bodily changes if they are present.

2. During the help-seeking interval, they perceive a reason to seek health care (due to symptoms, awareness of a recommendation to be screened for cancer, or any other reason) and take action to seek it.

3. The diagnostic interval begins at patients' first interaction with the health care system and ends with their diagnosis.

There are potential sources of diagnostic delay during each of these intervals. Delays that occur during the self-appraisal and help-seeking intervals are considered pre-health system delays, while those during the diagnostic interval are considered health system delays. Both pre-health system and health system delays may be influenced by contributing factors, such as patients' health-related knowledge and prior experiences with the health care system; provider and system factors, such as timely availability of appointments and insurance coverage; and disease-related factors, such the type of cancer, its presentation (e.g., nature of symptoms, if any), its speed of progression, and the availability of recommended screening tests.

Various methods have been used to assess timeliness of cancer diagnosis, including chart reviews and analysis of registry or claims data. Patient reports of diagnostic care experiences have the potential to complement these methods, helping to identify aspects of diagnostic timeliness that are ripe for quality improvement (Graber et al., 2025). Standardized surveys are the predominant method of collecting patient reports for use in measurement and improvement.

Patient surveys are widely used in the United States to derive patient-reported measures of care experiences and outcomes (Anhang Price et al., 2014). However, to date, there have been no patient-reported measures of timeliness of cancer diagnosis (or diagnosis of any other condition). The overarching goal of this work was to derive such measures; to do so, we developed and tested surveys that assess patient experiences of cancer diagnosis and analyzed survey response data to identify and describe the properties of survey-based quality measures.

Project Overview

In this project we aimed to develop patient-reported quality measures that assess timeliness of cancer diagnosis. To do so, we drafted and field-tested surveys that allow patients to report on their experiences throughout the cancer diagnostic process. In Phase 1 of the project (August 2022–December 2024), we focused on patient assessments of care during the diagnostic interval. To inform this effort, we conducted an environmental scan, including a literature review and interviews with subject matter experts (SMEs); convened a Technical Expert/Clinical User/Patient Panel (TECUPP); drafted a survey instrument and refined it based on the findings of cognitive interviews with individuals from across the United States who had recently been diagnosed with cancer; and field-tested the survey among patients recently diagnosed with cancer who were receiving care at Cedars-Sinai Health System in Los Angeles. We analyzed field test data to derive seven measures of timely diagnosis of cancer: the Patient-Reported Outcome Measures of Timely Experience of Diagnosis (PROMOTE-Dx) for Cancer.

In Phase 2 (January 2024–December 2025), we expanded the Phase 1 survey to include factors affecting diagnostic timeliness in the pre-health system intervals. To identify new survey content addressing these factors, we once again sought insights from the literature and SMEs, members of the TECUPP, refined survey questions and flow via cognitive interviews, and field-tested the resulting survey among a new sample of Cedars-Sinai patients who were recently diagnosed with cancer.

Environmental Scan

To inform field test survey content and administration procedures, we conducted an environmental scan consisting of a literature review and interviews with subject matter experts (SMEs).

To identify approaches to asking patients about timeliness of diagnosis, potential content areas, and specific survey questions for potential inclusion in our survey, we reviewed the peer-reviewed and grey literature. Specifically, we searched peer-reviewed literature in PubMed using terms related to cancer, diagnosis, and delay, and we reviewed existing patient experience surveys from sources such as the Agency for Healthcare Research and Quality and the United Kingdom National Health Service. From this review, we identified five main types of patient-reported questions—perceptions of delay, system or process issues, and communication and recalling time intervals and number of visits—and refined these into specific factors such as access to care, medical history, physical exams, tests and referrals, and communication and respect. We also identified pre-health system factors contributing to delays, including lack of knowledge, fear, financial barriers, and competing priorities. Using these findings, we created a survey question bank

We conducted interviews with 21 SMEs in fall 2022 to inform the Phase 1 field test and five additional interviews in spring 2024 to inform Phase 2. Interviewees were clinicians, researchers, cancer survivors or family caregivers, and quality measurement experts. The goal of the Phase 1 interviews was to identify key health system factors influencing timeliness of cancer diagnosis; gain expert input into the degree to which patient reports can contribute to understanding of timeliness, including benefits and challenges of collecting patient reports; and inform survey administration protocols, such as the appropriate lookback period and use of proxy respondents for those who may not be able to respond for themselves. The goal of the Phase 2 interviews was to identify common “help-seeking” or pre-health system factors that contribute to delays in cancer screening or seeking help for symptoms or another condition that may lead to an incidental finding.

Phase 1 interviewees reported that certain populations were more likely to experience delayed diagnoses due to limited access, communication barriers, and clinician bias. Interviewees identified failures in care coordination and communication as key contributors to delays and noted that patient perceptions of delay may not always align with clinicians' perspectives. Phase 2 interviewees identified factors contributing to pre-health system delays, such as uncertainty about symptoms, lack of awareness of screening recommendations, logistical challenges, mistrust, and fear as important barriers to seeking care.

Technical Expert/Clinical User/Patient Panel

We convened a 14-member TECUPP, including seven patients and caregivers with diverse lived experiences and seven clinicians and researchers representing primary care, specialty care, emergency medicine, and diagnostic quality research. We asked members to evaluate potential survey content areas (“domains”) based on their importance to timely diagnosis, applicability across cancer types, observability by patients, and actionability by health systems. We also polled members on the relative importance of specific factors within each domain and refined draft survey questions based on their feedback.

In these discussions, we found that TECUPP members prioritized the systems and processes and communication domains over those about time intervals or number of visits, which they viewed as less precise or meaningful in the context of the United States health care system. Members identified tests and referrals, communication and respect, access to care, and explanation and next steps as the most important and actionable systems, process, and communication factors for understanding diagnostic timeliness. They emphasized that patients can reliably report on communication and access issues and that health care providers can act on findings related to tests and referrals. For the pre-health system intervals, TECUPP members agreed that lack of knowledge, information from non-medical sources, fear or mistrust, financial barriers, and competing priorities were key contributors to pre-health system delays. They also suggested including survey questions about factors that both hinder and facilitate care-seeking, such as transportation, insurance, and advice from others. Based on these insights, we incorporated questions addressing these factors into the Phase 1 and Phase 2 field test surveys.

Drafting and Refining Survey Instruments for Field Testing

We drafted survey questions based on priorities identified by the environmental scan and TECUPP, adapting questions from existing validated instruments. We developed and refined surveys for Phase 1 and Phase 2 field tests through iterative rounds of cognitive interviews with a total of 35 cancer survivors representing diverse cancer types and educational and sociodemographic backgrounds.

Based on input from Phase 1 participants, we used the direct terms “diagnosed” and “diagnosis” rather than “being told you had cancer” throughout the survey; asked respondents to provide specific dates (i.e., of the first visit or test leading to cancer diagnosis and cancer diagnosis); described the diagnostic interval as “the time leading up to your cancer diagnosis,” visually depicting with a bracket that this time period starts with the date that the “very first” visit or test occurred leading to the cancer diagnosis and ends with the date of diagnosis; and placed a question regarding perceived timeliness of cancer diagnosis immediately after the questions about dates, to anchor respondents to the time interval of interest.

In response to feedback from Phase 2 participants, we organized the survey so that questions about symptoms were posed earlier and separately from questions about screening or incidental findings; populated the respondent's reported date of first visit into subsequent questions to help clarify the time interval of interest; and changed the response options on a number of questions.

Survey Field Tests

To evaluate survey administration methods, response rates, and measurement properties, we conducted two field tests of draft surveys among patients in the Cedars-Sinai Health System who were recently diagnosed with cancer.

Methods

In the Phase 1 field test (September 2023), patients with a recent cancer diagnosis (3–9 months prior to survey sampling) were identified through electronic health records and invited to complete an online survey, with mailed follow-ups for nonrespondents. In the Phase 2 field test (April 2025), to ensure sufficient sample size, eligibility expanded to include patients diagnosed in the three years prior to survey sampling, and surveys were administered exclusively online. Table 1 describes the eligibility criteria and administration procedures for each field test.

Table 1.

Phase 1 and Phase 2 Field Test Eligibility Criteria and Administration Procedures

	Phase 1 Field Test	Phase 2 Field Test
Inclusion Criteria	>1 cancer ICD-10 encounter diagnosis code in last 3–9 months (Dec 2022 to Jun 2023) Age 18+	>1 cancer ICD-10 encounter diagnosis code in last 3 months to 3 years (Jan 2022 to Dec 2024) Age 18+
Exclusion Criteria	Any encounters with cancer ICD-10 code prior to Dec 2022 Common skin cancers (C44) Cancer-like syndromes (C96) D-codes (in situ/stage 0 cancers)	Any encounters with cancer ICD-10 code prior to Jan 2022 Common skin cancers (C44) Cancer-like syndromes (C96) D-codes (in situ/stage 0 cancers)
Mode of administration	Email/web (2 emails), mail/paper (2 mailings)	Email/web (2 emails)
Incentive	None	$50 gift card if HIPAA waiver completed

Both the Phase 1 and Phase 2 field test survey instruments included questions regarding health system factors contributing to diagnostic delay, including aspects of communication, timeliness of appointments, and getting tests and results. In addition, the Phase 2 instrument included questions about factors contributing to pre-health system delays, including motivators for, demotivators for, facilitators of, and barriers to seeking health care prior to cancer diagnoses.

Using field test survey response data from both phases, we considered two questions as overall assessments of care during the diagnostic interval (overall rating of care during diagnostic interval and perceived timeliness of cancer diagnosis); for the remaining 10 evaluative questions about health care experiences, we conducted confirmatory factor analysis to evaluate the factorial structure (i.e., identify which questions could be grouped together to assess specific content domains of interest), and assessed the properties of the proposed groupings of questions (hereafter, composite measures) by calculating internal reliability of using coefficient Omega, relationship/overlap between the measures using Pearson correlations, and construct validity using linear and logistic regressions in which measures predicted several outcomes, including overall rating of care during diagnostic interval, perceived timeliness of cancer diagnosis, and diagnostic interval of less than or equal to 30 days (O'Hanlon et al., 2025).

Results

Using Phase 1 field test survey response data, we derived seven quality measures that assess health system factors, which we refer to as PROMOTE-Dx for Cancer:

• Two single-question overall assessments of care: Overall Rating of Care During Diagnostic Interval and Perceived Timeliness of Cancer Diagnosis

• Five measures assessing aspects of care quality during the cancer diagnostic interval, including:

  • Two single-question measures: Getting Timely Usual Care and Getting Timely Specialty Care

  • Three multi-question composite measures: Communication with Usual Care Professionals, Communication with Specialists, and Getting Tests and Results.

Using factor analysis, we found that the questions within the three multi-question composite measures strongly reflect the main themes of those measures. The five measures assessing aspects of care quality were moderately correlated, indicating that they assess related but distinct aspects of care. Results were similar for Phase 2.

Across the two field tests, comparable proportions of respondents indicated that they were diagnosed with cancer within 30 days of seeking care (i.e., a timely diagnostic interval: 47% in Phase 1, 41% in Phase 2) and that their health care providers could have diagnosed their cancer sooner (i.e., a potential diagnostic delay: 28% in Phase 1, 30% in Phase 2).

The most commonly reported motivators were going for a regular check-up (37.1% of respondents) and thinking I might be sick (36.3%) and most common facilitators were ease of getting a test or making an appointment (54.4%) and benefits or insurance (49.4%). More than two-thirds of respondents reported no demotivators or barriers; the most common was trouble scheduling an appointment (16.0%).

Guidance for Future Use

Sampling and Eligibility Criteria

We designed the surveys and PROMOTE-Dx for Cancer for use in health systems, health plans, or other types of entities that are accountable for all care delivered to a patient. This reflects the nature of the diagnostic process, which typically relies on care from (and coordination between) multiple providers or practices that are collectively rather than individually responsible for timely diagnosis. In addition, sampling at the level of the health system or health plan, rather than an individual clinician, would be expected to result in a larger number of completed surveys available for analysis, thereby improving statistical power.

To be eligible for the survey, we recommend that patients be age 18 or older and have a recent new cancer diagnosis (i.e., not a recurrence). In EHR, this may be defined as one or more cancer ICD-10 encounter diagnosis codes in the last 3 to 36 months, and no cancer codes in the preceding five years. If sample size allows, or if an organization is interested in focusing on diagnostic experiences during a specific timeframe within the last 36 months, the window may be shortened (e.g., 3–9 months).

To maximize the number of completed surveys for analysis, the survey is designed to assess the experiences of patients with all types of cancer. However, if sample size permits, users may choose to administer the survey to a cohort or stratified sample of patients with specific cancer types of interest (e.g., those diagnosed without a cancer screening test, solely breast cancer, etc.).

Selecting a Survey Version

Organizations may choose to administer a version of the survey that assesses both pre-health system and health system factors, tailoring questions and response options to their specific context. Alternatively, organizations may administer a 20-question, short-form version of the survey that allows them to calculate the seven PROMOTE-Dx for Cancer health system measures only (Table 2).

Table 2.

PROMOTE-Dx for Cancer Survey Questions and Response Options

Measure	Question(s)	Response Options
Getting Timely Usual Care (1 question)	From your first visit up until the day you got your diagnosis, were you able to see your usual health care professional(s) as soon as you needed?	Yes, definitely; Yes, somewhat; No; Don't know/Not sure
Communication with Usual Care Professionals (3 questions)	From your first visit or test that led to your cancer diagnosis up until the day you got your diagnosis, did your usual health care professional(s) listen carefully to you? take your concerns seriously? treat you with dignity and respect?	Yes, definitely; Yes, somewhat; No; Don't know/Not sure
Getting Timely Specialty Care (1 question)	Were you able to see the specialist(s) who helped diagnose your cancer as soon as you needed?	Yes, definitely; Yes, somewhat; No; Don't know/Not sure
Communication with Specialists (3 questions)	Did the specialist(s) who helped diagnose your cancer: listen carefully to you? take your concerns seriously? treat you with dignity and respect?	Yes, definitely; Yes, somewhat; No; Don't know/Not sure
Getting Tests and Results (2 questions)	From your first visit up until the day you got your diagnosis: how often was it easy to get the tests, x-rays, or scans you needed? when someone ordered a test, x-ray, or scan for you, how often did someone follow up to give you those results?	Never; Sometimes; Usually; Always; Don't know/Not sure
Overall Rating of Care During Diagnostic Interval (1 question)	Using any number from 0 to 10, where 0 is the worst care possible and 10 is the best care possible, what number would you use to rate the care you received from all your health care professionals from your first visit or test up until the day you got your diagnosis?	0 – 10
Perceived Timeliness of Cancer Diagnosis (1 question)	Do you think your health care professionals could have diagnosed your cancer much sooner than they did?	Yes; No; Don't know/not sure

Survey Administration

The version of the survey that includes pre-health system factors is complex, as it needs to orient respondents to specific dates and time intervals, and has different pathways of questions depending on initial cancer presentation. Given this, modes of administration that can be programmed to adapt to respondents' prior answers (e.g., web) is preferred to mail surveys.

The short-form survey addressing solely health system factors may be administered by mail, web, or phone.

Conclusions

Patient surveys can provide important information about cancer diagnosis experiences that can only be reported by patients, and that complement information available from other data sources. The surveys can be used to derive internally reliable and valid PROMOTE-Dx for Cancer for use in quality improvement.

Currently, the survey and the PROMOTE-Dx derived from it are designed to inform quality improvement. Further testing and validation of the survey among a broader range of organizations is needed to validate the measures for other uses, such as performance comparisons or public reporting.

This research was sponsored by the Gordon and Betty Moore Foundation and conducted in the Health Services and Outcomes Program of RAND Health.