Testing of a Tool for Prostate Cancer Screening Discussions in Primary Care

Background As prostate cancer (PCa) screening decisions often occur in outpatient primary care, a brief tool to help the PCa screening conversation in busy clinic settings is needed. Methods A previously created 9-item tool to aid PCa screening discussions was tested in five diverse primary care clinics. Fifteen providers were recruited to use the tool for 4 weeks, and the tool was revised based upon feedback. The providers then used the tool with a convenience sample of patients during routine clinic visits. Pre- and post-visit surveys were administered to assess patients’ knowledge of the option to be screened for PCa and of specific factors to consider in the decision. McNemar’s and Stuart–Maxwell tests were used to compare pre-and post-survey responses. Results 14 of 15 providers completed feedback surveys and had positive responses to the tool. All 15 providers then tested the tool on 95 men aged 40–69 at the five clinics with 2–10 patients each. The proportion of patients who strongly agreed that they had the option to choose to screen for PCa increased from 57 to 72% (p = 0.018) from the pre- to post-survey, that there are factors in the personal or family history that may affect PCa risk from 34 to 47% (p = 0.012), and that their opinions about possible side effects of treatment for PCa should be considered in the decision from 47 to 61% (p = 0.009). Conclusion A brief conversation tool for the PCa screening discussion was well received in busy primary-care settings and improved patients’ knowledge about the screening decision.

Methods: A previously created 9-item tool to aid PCa screening discussions was tested in five diverse primary care clinics. Fifteen providers were recruited to use the tool for 4 weeks, and the tool was revised based upon feedback. The providers then used the tool with a convenience sample of patients during routine clinic visits. Pre-and post-visit surveys were administered to assess patients' knowledge of the option to be screened for PCa and of specific factors to consider in the decision. McNemar's and Stuart-Maxwell tests were used to compare pre-and post-survey responses.
results: 14 of 15 providers completed feedback surveys and had positive responses to the tool. All 15 providers then tested the tool on 95 men aged 40-69 at the five clinics with 2-10 patients each. The proportion of patients who strongly agreed that they had the option to choose to screen for PCa increased from 57 to 72% (p = 0.018) from the pre-to post-survey, that there are factors in the personal or family history that may affect PCa risk from 34 to 47% (p = 0.012), and that their opinions about possible side effects of treatment for PCa should be considered in the decision from 47 to 61% (p = 0.009).
conclusion: A brief conversation tool for the PCa screening discussion was well received in busy primary-care settings and improved patients' knowledge about the screening decision.
Keywords: prostate cancer, cancer screening, shared decision-making, prostate-specific antigen, clinical decision-making inTrODUcTiOn The current US Preventive Services Task Force (USPSTF) Prostate Cancer (PCa) Screening Recommendations for men aged 55-69 suggest individualized decision-making for the PCa screening decision after a discussion with a clinician of benefits, harms, and consideration of a patient's values and preferences (1). For younger men, the previous American Urological Association guideline also suggests individualized screening decisions for men aged 40-54 at increased risk with a family history of PCa or for African-American men (2). Controversy remains regarding optimal PCa screening strategies (3) and the ability to engage in these conversations in primary care settings (4) where many PCa screening conversations may occur. We previously created a 9-item brief tool for PCa screening conversations (5). The goal of the tool is to aid PCa screening discussions where the patient and provider can discuss the risks and benefits of PCa screening given a patient's individual risk factor, health status, and preferences. We tested this tool with primary care providers and patients to determine its ease of use for providers and patients and to describe the responses in primary care settings.

MaTerials anD MeThODs
Each item on the PCa conversation tool is scored from 0 to 3 (minimum score 0, maximum score 27) with higher scores suggesting PCa screening may be beneficial. We tested the tool in three steps in five diverse urban and suburban primary care clinics. First, 15 primary care providers were recruited using an electronic mail announcement about the study to agree to use the tool for 4 weeks from June to July 2017 with up to 10 patients each. At the end of the 4 weeks, to assess provider experience with the tool, the validated Perceived Usefulness (6-item) and Perceived Ease of Use (6-item) scales (6) were sent electronically to providers through REDCap (7). The 12 questions on these two scales have seven options ranging from extremely likely to extremely unlikely. The conversation tool was revised based upon feedback and a final version was created (Appendix S1 in Supplementary Material). In the last step, the same group of 15 providers was asked to use the tool with a convenience sample of patients during routine clinic visits from September to December 2017. A member of the research team reviewed provider schedules for eligible male patients aged 40-69 who had no prior history of PCa. If the provider and patient agreed to participate in the study at the time of the appointment, a 4-item pre-visit survey was administered to the patient by a research assistant. The provider then used the conversation tool with the patient. Previously created reference information that was available through the electronic medical record to all of the participating primary care providers regarding PCa screening, thus part of usual care, was made available to the provider in hard copy to use during the visit. After the visit, a 5-item post-visit survey was administered. These surveys had questions related to patients' knowledge that they have an option to be screened for PCa and of specific factors to be considered in the decision. Each question was scored on a 5-point Likert scale with a range of strongly disagree (1) to strongly agree (5) Table 1. Further analysis of the proportion of patients who strongly agreed that they had the option to choose to screen for PCa increased from 57 to 72% (p = 0.018) from the pre-to post-survey, that there are factors in the personal or family history that may affect risk for PCa from 34 to 47% (p = 0.012), and that their opinions about possible side effects of treatment for PCa should be considered in the decision from 47 to 61% (p = 0.009). When selecting from a list, the three most important factors in the PCa screening decision, the most frequent responses were age (65%), family history of PCa (42%), and concern about developing PCa (36%). Concern about sex life and leakage of urine were among the three most important factors for 10 and 18% of patients, respectively ( Table 1). For the question of choosing the three most important factors in the PCa screening decision, we also compared responses of patients age <55 to those My concern about leakage of urine 17 (18%) My race or ethnicity 14 (15%) My concern about sex life 9 (10%) option to screen for PCa, that there were factors in their personal and family history that may affect their risk for PCa, and that their opinions about possible side effects of treatment for PCa should be considered in the screening decision. As very few patients scored very low or high on the tool, responses did not readily direct the patient to a decision that screening would or would not be preferred. However, the tool effectively introduced the screening conversation. Discussions about PCa screening may not be of high quality (8) and cancer screening guidelines may not always provide the optimal level of information to aid these discussions (9). Allowing patients to understand that they have an option to make a decision is a key component of operationalizing shared decision-making in a clinical setting (10). Decision aids for PCa screening can improve patient knowledge and involvement in decision making (11) and the interaction with a provider in addition to decision aid use alone may improve patient understanding of the PCa screening decision (12). The use of our tool prompted providers to cover the key domains that are relevant to a PCa screening decision and identify those men who could benefit from more extensive shared decision-making conversations if they scored in the mid-range on the tool.
Interestingly, when asked post-visit, patients most frequently chose their age, family history, and concern about developing PCa as important to the screening decision. Concern about leakage of urine or sex life were cited less frequently, although these are the most common potential downstream effects of PCa treatment that are often discussed. Our findings highlight that understanding and addressing patients' anxiety about the PCa diagnosis should be an important component of the conversation, especially as related to ability to accept no immediate treatment for low-risk disease (13), given the option of active surveillance in this situation. While providers had access to reference information regarding population risks and benefits of PCa screening, decision aids providing individualized risk prediction of PCa screening outcomes, including the likelihood of additional downstream testing, may further improve the quality of the screening discussions, as patients often do not understand harms of screening tests (14) but rather focus on benefits.
A limitation of our study is that 40% of patients enrolled were in the younger age group of 40-54, thus our summary findings may not fully reflect the views of men in the age group (55-69) that the current USPSTF recommends for individualized decision-making for PCa screening. However, we believe our findings across these age groups remain relevant to practicing clinicians.

cOnclUsiOn
We demonstrated the usefulness of a brief PCa conversation tool in primary care settings to improve PCa screening conversations and to identify the need for further shared decisionmaking around the PCa screening decision. Future work will focus on the implementation of the PCa tool in additional settings and on the outcome of the screening decision after use of the tool.
≥55 years and found no significant differences in frequency of responses in the two age groups (results not shown).

DiscUssiOn
Our study demonstrated the feasibility of using a PCa screening conversation tool in busy primary-care settings to address important aspects of the PCa screening discussion. Providers found the tool easy to use and after the tool was used during the visit, patients were more likely to strongly agree that they had an