PDF
Citation
Print
INTRODUCTION
Prostate cancer is a common cancer among men, and its incidence is rapidly increasing in Korea, with an estimated ranking of #1 among cancer cases in men by 2022.1 However, the counterpart trends in many countries have been decreasing or stabilizing for the past five years, especially in high-income countries.2 This is likely due to a reduction in the use of prostate-specific antigen (PSA)-based screening. For example, routine PSA-based screening and prostate cancer incidence rates in the US sharply declined after the US Preventive Service Task Force (USPSTF) recommended against routine PSA testing in 2012.3 This is in contrast to the rapid increases of prostate cancer incidence in the late 1980s to early 1990s in the US.4
In healthy individuals, PSA-based routine screening for prostate cancer is not recommended, given the conflicting results with mortality reduction.56 Instead, many professional organizations highlight informed decision making for prostate cancer screening among certain ages. The USPSTF revised its guidelines to recommend informed decision making for men aged 55 to 69 years.7 The American Urological Association (AUA) recommends shared decision-making (SDM) for PSA-based screening for men aged 55 to 69 years.8 The Japanese Urological Association recommends that PSA-based screening and its risks and benefits should be offered to at-risk men.9 The Korean Urological Oncology Society promotes prostate cancer screening in high-risk Korean men 50 years and older.10
Currently in Korea, the National Cancer Screening Program does not include prostate cancer,11 recommending screening only for the stomach, colon, lung, liver, breast, and cervical cancer. Awareness of appropriate cancer screening is low among citizens, which affects the rates of screening. In Korea, the peak age-specific incidence of prostate cancer is 75–79 years.12 However, for example, Koreans aged 48 are notified by the National Health Insurance Service (NHIS) to participate in stomach cancer screening and may assume that they also need prostate cancer screening. In addition, PSA testing is easy to perform using only a blood test, and many health screening providers offer PSA tests without the appropriate SDM process.
A decision aid (DA) is an intervention that supports patients in decision making about available options.13 DAs enhance decision quality by augmenting patients’ knowledge about available options and their associated risks and benefits, and by assisting in the clarification of the alignment between decisions and personal values.14 In Korea, to our knowledge, there is no DA available for prostate cancer screening. Hence, we aimed to develop a web-based DA for prostate cancer screening that is tailored to individual risk profiles and to evaluate its potential effectiveness in improving informed decision making.
METHODS
The International Patient Decision Aid Standards (IPDAS) development process model (Fig. 1) was used in the development of the present DA. Here, we focused on the five steps of the process: 1) scope, 2) design, 3) prototype, 4) alpha testing, and 5) beta testing. The first three comprise the development phase, and the last two are the evaluation phase.
Scope
The scope of an IPDAS model involves determination of the target audience and definition of the purpose of the DA. To decide if healthy men should undergo screening for prostate cancer, information on screening and individualized estimated risk of prostate cancer is needed. Based on that, we aimed to develop a DA to provide knowledge, personalized risk and benefit assessment, and values for decision making that target the general male population.
Design
The design process in an IPDAS model identifies patient needs, elicits views of health professionals on patient needs, and reviews and synthesizes evidence. For this, we invited multidisciplinary experts (urologist, psychologist, and nurses specialized in cancer education) to participate in semi-structured face-to-face interviews and elicited expert opinions on the unmet needs of prostate cancer screening information and the optimal methods to provide such information.
Research teams (family physicians in academic hospitals) have reviewed current international guidelines78915 and DAs of prostate cancer screening.1617181920 Five such DAs in formats of PDF, webpage, video clips, or software program were identified and reviewed (Supplementary Table 1). All but one DA targeted the general male population considering prostate cancer screening. The fifth DA specified the target age group of 50 years or older.19 None of the DAs provided individual risk of prostate cancer, and all included brief information on PSA as a blood test. Benefits of prostate cancer screening was based on the results from the European Randomised Study of Screening for Prostate Cancer (ERSPC),6 demonstrating 1 mortality reduction among 1,000 people who underwent screening for prostate cancer more than two times during 13 years of follow-up. In addition, detection of prostate cancer in early stages was included in the benefits of screening. The risks of PSA-based screening include false positives, false negatives, complications resulting from additional invasive examinations, overdiagnosis, and possible overtreatment. The risks and benefits were described by infographics of pictograms for four DAs16181920 and as text for the fifth.17 Three DAs provided personal values and preferences toward prostate cancer screening and enquired about the final decision on screening.161819 Personal values were considered by three DAs,161819 and among them, “Yes or No” required age, sex, race, and area of residence to fill in, and 10 questions that asked scientific evidence in screening.19 By integrating these components, we developed a prototype DA.
Drafting of a DA prototype
A prototype is a draft DA that includes a storyboard and script. The prototype of the DA for prostate cancer screening was developed through a collaborative effort by the research team, which included family medicine physicians, and multidisciplinary experts. The team agreed on an open-access, web-based format for the final prototype. The prototype comprises two parts, namely risk prediction and decision making. To assist in decision making, an individual risk prediction model was developed and validated,21 and the risk was presented as an absolute value (e.g., 9-year risk = X%) and relative risk (e.g., 1.5 times higher than men of the same age). The focus was on highlighting the negative aspects of screening in those with low prostate cancer risk rather than instilling fear of potential risk.
The decision making process involved five components: knowledge, risk and benefit, cost, personal value, and decision making. The knowledge component aimed to provide general information on prostate cancer and screening through the use of animated video clips. The second component addressed the risks and benefits of screening, including mortality reduction, false positives, false negatives, overdiagnosis, and complications from subsequent invasive examinations. The third component focused on cost considerations. To determine personal values regarding early detection of prostate cancer, false positives, and adverse events, we used a 4-point Likert scale. Finally, the decision to screen was made based on the information provided in the previous components.
Alpha testing (usability test)
A usability test was performed on the prototype screens among healthy individuals. Initially, 12 individuals (7 men, 5 women; mean age: 60.0 ± 12.6 years) were recruited via advertisement to provide written feedback on the revised prototype. The participants evaluated the prototype’s ease of understanding, difficulty of terms, helpfulness, and length using a four-item Likert-type scale and provided explanations for each answer. Moreover, free-text comments about the DA were solicited. Of the 12 participants, 3 reported that the DA was very easy to comprehend, and 9 found it easy to understand. Although all participants agreed that the terms used in the DA were not difficult, some participants expressed difficulty in understanding some medical terms in Korean, such as “nodule” and “follow-up.” Six participants found the content of the DA to be very helpful, five found it helpful, and one person reported that it was not helpful. Although four individuals considered the length of the DA to be long, the majority of the participants8 found it to be of an acceptable length. In the free-text comments, some participants suggested that the DA could discourage prostate cancer screening due to the possibility of overdiagnosis.
The research team agreed to replace some medical terms with plain language to enhance understanding and balance the information of pros and cons regarding prostate cancer screening. To further customize the prototype, advisory meetings were held with multidisciplinary experts. The information was visually presented as absolute risk in a bar graph, risk compared to those of the same age, and risk of men of the same age with no risk factors (see Fig. 2).
Fig. 2
Overview of decision aid. (A) Information for prostate cancer risk prediction, (B) Predicted risk of prostate cancer, (C) Epidemiology of prostate cancer, (D) Symptom of prostate cancer, (E) PSA test, (F) Benefits of prostate cancer screening, (G) Risks of prostate cancer screening, (H) Risks of PSA-based screening: complications from biopsy and surgery.
PSA = prostate-specific antigen.
Beta testing (pilot evaluation study)
The eligibility criteria for our pilot evaluation study were set as healthy men aged between 40 and 79, drawn from the general population, and without a personal or familial history of prostate cancer. A history of other cancers, or a family history of such, was not considered exclusionary. As the primary objective of pilot studies like ours is to establish research design feasibility, identify potential procedural complications, assess effect size, and ascertain variability in outcome measures, our sample size decision was not informed by traditional statistical power calculations. For our purposes, a sample size of 100 was deemed adequate—large enough to enable meaningful analysis and interpretation, yet small enough to avoid unnecessary resource consumption that could be more beneficially directed towards the primary study.
A total of 101 men aged between 40 and 79 years were recruited to participate in this study, using purposive sampling to recruit 25 participants in each 10-year age group who were fluent in both spoken and written Korean. Recruitment was performed within the general population in four districts of the Seoul metropolitan area: Kangbuk, Kangnam, Kangseo, and Kangdong. Our strategically located recruitment booths in these regions optimized participant engagement. Trained researchers explained the purpose of the study to potential participants and obtained informed consent from each of them.
Computer-assisted personal interviews were conducted to assess participants’ knowledge, attitude, intention to screen, decisional conflict, and usefulness of the decision aid. Pre- and post-comparisons were made for the first three domains, while the latter two were measured only after participants had used the DA.
Knowledge: To evaluate participants’ knowledge of prostate cancer screening, we developed a 10-item questionnaire consisting of yes/no questions based on the information covered in the DA.
Attitude: Participants’ attitudes towards prostate cancer screening were measured with a single item: “Do you think you should undergo prostate cancer screening?”. Participants used a 5-point Likert scale ranging from 1 (not at all) to 5 (very much) to indicate their level of agreement.
Intention to screen: Participants were asked about their intention to screen for prostate cancer and responded with yes, no, or don’t know.
Decisional conflict and usefulness: Each component was evaluated by the SURE tool® (four items measured on a 2-point Likert scale) for decisional conflict22 and the Preparation for Decision Making Scale®(10 items measured on a 5-point Likert scale) for usefulness.23
In addition, demographic information such as age, sex, area of residence, marital status, education level, household income, past medical history, family history of cancer, and prostate cancer screening in the past two years were also collected. Lastly, participants were asked to indicate whether they found the DA helpful in making a decision regarding prostate cancer screening. Further details regarding the survey questionnaire questions can be found in the Supplementary Table 1.
Statistical analysis
To test for normality, the Kolmogorov-Smirnov test was applied to continuous outcomes such as knowledge and attitudes. Paired t-tests were used for continuous outcomes with normal distribution, while Wilcoxon signed-rank tests were used for continuous outcomes without normal distribution, to evaluate differences before and after using the DA. To examine the difference in ternary outcomes (e.g., intention to screen: yes, no, unsure) before and after using the DA, the Stuart-Maxwell test was utilized. Moreover, subgroup analysis was performed based on participants’ history of prostate cancer screening within the past two years. All statistical analyses were conducted using STATA/MP 14.0.
RESULTS
Beta testing (pre-post study)
A total of 101 participants with a median age of 60 years (interquartile range 50–69 years) were included in the beta testing phase. The majority (84%) had not undergone prostate cancer screening in the past two years, and 26% reported a family history of cancer, while 1% had a personal history of cancer. Further details are presented in Table 1.
Table 1
Baseline characteristics of the study population (N = 101)
Effectiveness of the DA
Knowledge and attitude
The use of the DA led to an increase in participants’ knowledge of prostate cancer screening (6.85 ± 1.03 before and 7.57 ± 1.25 after, P < 0.001) (Table 2). However, it did not change their attitude toward prostate cancer screening (Wilcoxon signed-rank test P = 0.564; Fig. 3). Before using the DA, only 14.9% of participants believed that prostate cancer screening was not needed or very not needed; this percentage increased to 24% after using the DA.
Table 2
Effectiveness of the decision aid
Data are expressed as percentage (%) or mean ± standard deviation.
Knowledge: On a 10-item yes/no-response questionnaire (score range 0–10).
Attitude: Negative attitudes were assessed by participant (%) answers to the question “Do you need prostate cancer screening?” as 1 (not at all) or 2 (a little); (1–5 Likert-type scale).
aPaired t-test.
bWilcoxon signed-rank test.
cStuart-Maxwell test.
dPercentage of participants with Sure tool® score 4 (lowest decisional conflict); (range 0–4) (details in Supplementary Table 3).
ePreparation for Decision Making Scale® score; (range 0–100) (details in Supplementary Table 4).
Intention to screen
After using the DA, 51.5% of men reported an intention not to screen, up from 27.7% before (Stuart-Maxwell P < 0.001; Fig. 4). Among 58 participants who initially aimed to undergo screening, 43 (mean age 65.5 ± 10.1 years) retained their intention, while 13 (mean age 55.1 ± 8.2 years) changed their intention to not undergo screening. All 28 participants (mean age 54.3 ± 10.9 years) who initially had no intention of screening for prostate cancer reported the same intention after using the DA. Among 15 participants who were unsure before the DA, 11 (mean age 51.8 ± 8.9 years) reported no intention to undergo screening after using the DA.
Decisional conflict and perceived usefulness
After using the DA, 79 participants reported no decisional conflict (Fig. 5). The usefulness score of the DA was high (mean score [of 100] 77.35 ± 7.69), with 85% of men responding that the DA helped with decision making. In contrast, 12% of respondents answered that the DA seemed to recommend prostate cancer screening.
Subgroup analysis
In subgroup analysis according to prostate cancer screening within the past two years, knowledge of prostate cancer screening was increased in participants who did not undergo prostate cancer screening in the past two years (Screened/Non-screened) (P < 0.001). Such knowledge was not increased in the counterpart group (Yes) (Table 3). No significant difference of pre-post change in knowledge was found between the two subgroups (P = 0.057). While attitude toward prostate cancer screening was not changed in either group, intention to not undergo screening increased markedly in both groups: 31.8% to 58.8% (P < 0.001) for “No/don’t know” and 6.2% to 12.5% (P < 0.006) for “Yes.”
Table 3
Effectiveness of the decision aid according to prostate cancer screening
Data are expressed as percentage (%) or mean ± standard deviation.
Knowledge: On a 10-item yes/no-response questionnaire (score range 0–10).
aPaired t-test.
bT-test for difference of difference.
cNegative attitude for those not screened for prostate cancer; assessed by participants (%) answers to the question “Do you need prostate cancer screening?” as 1 (not at all) or 2 (not a lot); (1–5 Likert-type scale).
dPositive attitude for those screened for prostate cancer; assessed by participants (%) answers to the question “Do you need prostate cancer screening?” as 4 (a lot) or 5 (very much); (1–5 Likert-type scale).
eWilcoxon signed-rank test.
fStuart-Maxwell test.
gPercentage of participants with Sure tool® score 4 (lowest decisional conflict); (range 0–4).
hPreparation for Decision Making Scale® score; (range 0–100).
After using the DA, 76.5% and 87.5% participants in the screened and non-screened groups reported no decisional conflict. No difference in usefulness of DA was noted between the groups (P = 0.601).
DISCUSSION
The present study introduces the initial Korean DA for prostate cancer screening that aligns with the present guideline’s recommendation against routine screening. Notably, implementation of this DA resulted in improved informed decision making, as evidenced by increased participant knowledge and reduced intention to undergo screening. Participants also reported high perceived usefulness of the DA. These findings highlight the potential value of DAs in promoting informed decision making in the context of prostate cancer screening.
Remarkably, our DA demonstrated a decrease in screening intention. In contrast, previous pre-post studies exhibited mixed results on intention. Specifically, three studies reported a decrease in intention2425 or no effect,26 whereas one study reported an increase in screening intention.27 It is noteworthy that the latter study aimed to promote screening among African Americans, who have a relatively low screening rate despite a high risk of prostate cancer compared to other racial/ethnic groups.
We propose several explanations for our study’s findings. First, our DA significantly improved knowledge on prostate cancer screening, which is often the initial step toward changes in both intention and behavior for individuals.28 Notably, our DA led to a significant increase in the percentage of correct responses to questions on early prostate cancer treatment (30.7% to 70.3%), PSA testing methods (21.8% to 43.6%), and PSA overdiagnosis (71.3% to 91.1%) (Supplementary Table 2). Thus, the acquisition of information on the negative aspects of prostate cancer screening may have influenced participants' intention to undergo screening. However, it is imperative to acknowledge a notable reduction in the participants’ understanding of the limitations of the PSA test (27.7% to 18.8%), which does not conclusively detect all instances of prostate cancer, thus increasing the risk of false negatives. Additionally, there was a reduction in understanding regarding the impact of PSA tests on mortality reduction (21.8% to 16.8%). This was an unexpected result, potentially attributed to the detailed nature of the information relayed through our DA, which elucidated both the risks and benefits associated with PSA testing for prostate cancer screening. Given the scope of our inquiries, encompassing the decision to undergo PSA testing, the information was communicated through our DA in a manner that neither guaranteed complete detection of prostate cancer nor assured morality reduction. This possibly incited confusion among the participants, leading to a decrease in accurate response rates. Intriguingly, usability testing of our DA prototype, which examined factors such as ease of understanding, difficulty of terms, helpfulness, and length, revealed that all participants found it easy or very easy to comprehend, a discrepancy that calls for further exploration. To address these challenges, revisions to our DA to enhance understanding of prostate cancer screening are suggested, given the pilot nature of this study.
Second, our individualized risk calculator provided participants with personalized statistics on their individual prostate cancer risk, which may have corrected overestimations of incidence rates and led to changes in intention to undergo screening. Discrepancies often exist between perceived risk and actual estimated risk of prostate cancer. A previous survey of Korean men reported that the majority (85.9%) overestimate their lifetime incidence rate of prostate cancer, with an actual lifetime cumulative incidence of less than 10%.29 Our risk calculator demonstrated a 9-year prostate cancer incidence rate of 0.783%, 2.774%, and 4.132% for men of average risk in their 50s, 60s, and 70s, respectively.
Subgroup analyses revealed that our DA was notably more effective among individuals who had never undergone screening, exhibiting greater increases in knowledge and a higher intention not to undergo screening compared to the screened group. This unscreened group had a younger mean age of 57.4 ± 11.1 years and higher education levels than their screened counterparts with a mean age of 67.4 ± 9.3 years. The current prostate cancer screening guidelines target shared decision making for this age group globally.7891015 Therefore, our DA holds clinical significance by providing decision-making assistance, particularly for the younger age group.
The low level of decisional conflict combined with a high perceived level of preparation for decision making further validates the potential utility of our DA in facilitating SDM in real-world clinical settings. In Korea, cancer screening is frequently conducted without the benefit of physician counseling, and unlike screening guidelines for other cancers such as gastric or colorectal cancer, there is no universally accepted consensus or guideline for prostate cancer screening.11 In light of these circumstances, our freely accessible web-based DA is likely to be highly beneficial and has been made available on the website of the Korean Cancer Society (http://www.kcscancer.org/).
In conclusion, our web-based DA enhanced informed decision-making on prostate cancer screening, resulting in increased knowledge, decreased intention to screen, and high perceived usefulness. Our DA may have clinical implications for aiding decision making, particularly for the younger age group. Further larger scale studies to evaluate the effectiveness of DA on prostate cancer screening are warranted, such as randomized controlled trial.
XML Download