Trends in cancer screening rates among individuals with serious psychological distress: an analysis of data from 2007 to 2016 Japanese national surveys

Abstract

Background: 

The aim of the present study was to examine whether the increasing trend in cancer screening rates over time in Japan differs between individuals with and without serious psychological distress (SPD).

Methods: 

We analyzed a national representative cross-sectional dataset from the 2007, 2010, 2013, and 2016 Comprehensive Surveys of Living Conditions of Japan. We included people under the age of 69 years who met the national program criteria for each type of cancer screening (colorectal, gastric, and lung: n = 94,690, 94,957, and 94,751, respectively). SPD was defined as a score of ≥13 on the Kessler 6 scale. We calculated these cancer screening rates with 95% confidence intervals, stratified by presence or absence of SPD to determine whether cancer screening rates over time between 2007 and 2016 differ between individuals with and without SPD.

Results: 

Cancer screening rates in the non-SPD group increased steadily over time from 2007 to 2016 (colorectal, gastric, and lung: 31.4%–46%, 37.2%–44.8%, and 32.1%–51.7%, respectively). They also increased similarly over time in individuals with SPD (colorectal, gastric, and lung: 24.5%–36.3%, 30.2%–37.8%, and 24.8%–39.9%, respectively); however, the rates remained significantly lower than those of people without SPD in all survey years.

Conclusions: 

Between 2007 and 2016, cancer screening rates increased similarly over time in individuals with and without SPD. However, gaps in cancer screening in people with SPD remained unresolved.

Introduction

People with serious psychological distress (SPD), including mental illness, have been shown to have a higher mortality rate than those without it.^[1–4] Unfortunately, that mortality gap has been growing over time,^[1,3] indicating that further public efforts are needed to reduce it. Cancer is one of the common most causes of excess deaths in people with SPD.^[2] Cancer screening is important in reducing cancer mortality and previous studies have shown gaps in cancer screening for people with SPD.^[5–7] The low cancer screening rates among people with SPD may be owing to complex factors including patient, provider, and systemic factors.^[6,8] For example, patient factors may include low motivation, self-neglect, and socioeconomic disadvantages associated with SPD.^[8,9] In addition, associations have been found between people with SPD and limited health literacy levels.^[10]

Improving cancer screening rates is a global challenge and various strategies for increasing these rates have been proposed.^[11] Cancer screening rates remain lower in Japan than in other developed countries,^[12] which may be owing to health system factors in the country, such as the lack of a unified, nationwide, population-based cancer program and rarity of community-oriented general practitioners.^[13] Therefore, improving participation in cancer screening has been prioritized in Japan in particular. The Cancer Control Act was implemented in Japan in 2007 with the goal of increasing cancer screening rates to >50% within 5 years.^[14] Municipalities have implemented various strategies to achieve this goal, including sending personal invitations, personal visitations by community health workers, screening individuals in a clinical setting, and free screening.^[15] The results of the National survey indicate that cancer screening rates in Japan have been increasing in recent years^[16]; however, the target has not yet been achieved.

Additionally, it is unclear whether the gaps in cancer screening rates among people with SPD have resolved in recent years. Little is known regarding the time trends of the association between psychological distress and cancer screening rates worldwide. Previous studies investigating trends in gaps in cancer screening rates between people with and without mental disorder have shown inconsistent results, the gap variously reported as having widened or remained almost unchanged.^[17,18] Our previous study, using an anonymized data set (n = 93,730) from the 2010 Comprehensive Survey of Living Conditions (CSLC) conducted by the Ministry of Health, Labour, and Welfare (MHLW) of Japan, showed significant gaps in cancer screening rates among people with SPD in Japan in 2010.^[7] Specifically, people with SPD were less likely to participate in colorectal cancer screening (adjusted odds ratio [OR], .743; 95% confidence interval [CI], .638–.866), gastric cancer screening (OR, .823; 95% CI, .717–.946), and lung cancer screening (OR, .691; 95% CI, .592–.807).^[7] However, it is unclear whether the gaps are widening or narrowing over time.

The aim of the present study was to examine whether the trend of cancer screening rates increasing over time between 2007 and 2016 in Japan differs between individuals with and without SPD.

Materials and methods

Setting

In Japan, all residents are supposed to join the public health insurance program and individuals’ medical expenses are kept below specified limits according to their income and age. People can undergo cancer screening provided by municipalities or collective opportunistic cancer screening subsidized by insurers, which may either be free or include a low copayment. Local governments and workplaces have played major roles in encouraging cancer screening, whereas physicians have not been actively involved in recommending it because Japan has no general practitioner registration system. All residents who are eligible are encouraged to undergo cancer screenings using a leaflet or brochure distributed by the local government. However, the content of the leaflet/brochure is not individualized to local residents.

Data Source

The CSLC, a national cross-sectional survey, is conducted annually by the MHLW of Japan.^[19] Questions about health, household, and income are asked every 3 years. We used deidentified individual-level CSLC data sets of 2007, 2010, 2013, and 2016 with the approval of the MHLW. The sample of CSLC participants was randomly selected from approximately 5410 of 5530 districts included in the 2005/2010 National Census of Japan, covering approximately 290,000 households.^[20] The response rates for household and health questionnaires were 79.9% for 2007, 79.1% for 2010, 79.4% for 2013, and 77.5% for 2016. The sample of the income questionnaire survey comprised about 30,000 to 50,000 households in around 2000 unit areas selected by a stratified random sampling method from subareas of the above-mentioned 5410 of 5530 areas. The response rates for the income questionnaire were 64.8% for 2007, 72.6% for 2010, 72.5% for 2013, and 71.8% for 2016. In the present study, eligibility for analysis was having responded to all 3 questionnaires.

Participants

We included individuals aged 40 to 69 years for analyses of colorectal, gastric, and lung cancer screening. We excluded participants on the basis of the following criteria: in-hospital or in-social welfare facilities on the survey date (or this information was unknown); currently attending outpatient clinics for malignant neoplasm, pregnancy, or postpartum care; and not independently mobile.

Variables

Cancer screening

The main outcome variables were participation in colorectal, gastric, or lung cancer screening within 1 year. Up until 2016, the MHLW recommended the following cancer screening procedures: fecal occult blood testing for colorectal cancer screening, upper gastrointestinal radiography for gastric cancer screening, and chest radiographs for lung cancer screening for individuals aged 40 years and older every year. The CSLC questionnaires asked whether respondents had undergone these recommended screening tests as well as upper endoscopy for gastric cancer screening, which was not included in the MHLW list of recommended tests before 2016. As a side note, since 2016, gastric cancer screening by upper gastrointestinal radiography or upper endoscopy every 2 years has been recommended for individuals aged 50 years and older.

Breast and cervical cancer screening every 2 years are also recommended by the MHLW. As described in the next section, associations between cancer screening behavior in the past 2 years and SPD in the past month were outside of our area of interest because it is problematic to assume that SPD in the past month denotes SPD in the past 2 years.

Psychological distress

Since 2007, the Kessler 6 (K6) scale has been used to evaluate psychological distress in the CSLC. The K6 scale was developed as a brief screening scale for nonspecific psychological distress (PD) in adults and consists of 6 questions asking about symptoms of PD during the past 30 days.^[21,22] We classified individuals with scores of ≥13 (possible score 0–24) as having SPD.^[21]

Other variables

We included the categorical variables of age, sex, employment status, marital status, type of health insurance, household income, smoking status, presence or absence of current visits to outpatient clinics for physical illness, and independence of daily living as potentially confounding variables.

Ethical considerations

This study used deidentified individual-level CSLC data sets based on Article 33 of the Statistics Act and was approved by the MHLW. Our analysis using these data sets was conducted independently; thus, our statistical data may differ from those published by the MHLW.

Statistical Analysis

All analyses were conducted with SPSS version 22 (IBM, Tokyo, Japan). We obtained the background characteristics of the participants. We calculated colorectal, gastric, and lung cancer screening rates with 95% confidence intervals (CIs), stratified by presence or absence of SPD to examine whether cancer screening rates between 2007 and 2016 differed over time depending on SPD status. We calculated the odds ratios (ORs) with adjustment for all covariates and interaction between SPD and the survey year to examine whether changes in cancer screening rates differed according to SPD status.

Results

Eligibility criteria were met by 109,861 persons from the 4 data sets (2007, 2010, 2013, and 2016). Of these 109,861 persons, 10,553 (9.6%) had an unknown total K6 score and 7352 (6.7%), 7064 (6.4%), and 7292 (6.6%) respondents had unknown histories of colorectal, gastric, and lung cancer screening in the past year, respectively. We analyzed data of 94,690 individuals for colorectal cancer screening, 94,957 for gastric cancer screening, and 94,751 for lung cancer screening. Table 1 shows the characteristics of participants for whom each type of cancer screening was recommended in the 4 study years; some of these individuals had not actually undertaken the recommended screening. The prevalence of SPD was 3.6% in participants for whom any type of cancer screening was recommended.

Table 1: Characteristics of participants recommended for each type of cancer screening.

Figure 1A–C shows rates for each type of cancer screening with 95% CIs. The colorectal cancer screening rates among people without SPD increased steadily over time from 31.4% in 2007 to 46.0% in 2016. The colorectal cancer screening rates of people with SPD remained significantly lower than those of people without SPD in all survey years. However, the screening rates of those with SPD increased similarly over time from 24.5% in 2007 to 36.3% in 2016. As with colorectal cancer screening, gastric and lung cancer screening rates among people without SPD increased steadily over time from 37.2% to 44.8% and from 32.1% to 51.7%, respectively. These rates for people with SPD also remained significantly lower than those of people without SPD in all survey years; however, these rates increased over time from 30.2% to 37.8% and from 24.8% to 39.9%, respectively.

Figure 1: Trends in cancer screening rates among people with and without serious psychological distress. CI = confidence interval, SPD = serious psychological distress.

Table 2 provides the ORs for participation in each type of cancer screening according to SPD status. People with SPD were less likely to participate in each type of cancer screening than those without SPD (adjusted OR [95% CI] = .798 [.665–.958] for colorectal; .821 [.690–.976] for gastric; .772 [.643–.927] for lung cancer screening). The interaction terms between SPD and survey year were not significant for any of the 3 types of cancer screening.

Table 2: Adjusted odds ratios^∗ for participation in each type of cancer screening

In addition to the primary outcomes of the interaction terms between SPD and survey year, Table 2 also provides the ORs for participation in each type of cancer screening according to other demographic covariates. Sex, age groups, residential areas, marital status, health insurance, household income, smoking, current outpatient visit, and daily living independence level were significantly associated with participation in all 3 type of cancer screening.

Discussion

The increases in cancer screening rates over time between 2007 and 2016 in Japan did not differ significantly according to SPD status; that is, the gaps between people with and without SPD did not widen over time. However, our findings demonstrate that these gaps remained unchanged between 2007 and 2016. Recent public activities to promote cancer screening may have been equally effective in people with and without SPD, but not effective enough to eliminate the gaps.

There are few previous studies on the time trends of the association between psychological distress and cancer screening rates. In a previous Korean national database study that investigated trends in cancer screening rates in relation to disabilities, including mental disorder, over time,^[17,18] the gap in cervical cancer screening rates between people with and without mental disorders clearly widened.^[17] However, in that Korean study, the gap in colorectal cancer screening rates remained almost unchanged over time.^[18] Consistent with the results of the latter study, the present study showed that the gaps in colorectal, gastric, and lung cancer screening rates did not widen but also remained almost unchanged. However, we did not investigate time trends in the gap in cervical cancer screening rates in the present study because in Japan there has not been a significant gap in cervical cancer screening rates between people with and without SPD.^[7]

Our findings suggest that recent public activities to promote cancer screening have been insufficient to eliminate the disparities in cancer screening rates in Japan and indicate there is a need for policy makers and researchers to address these disparities. In Japan, common routes for making cancer screening recommendations are the mailing of personal invitations by municipalities and recommendations for cancer screening in the workplace. In addition, the national government has instructed family physicians to encourage their patients to participate in cancer screening.^[23] Reducing or eliminating the disparities in cancer screening rates between people with and without SPD will likely require mental health professionals, who may be the only health care providers with whom people with SPD interact, to play an important role in encouraging such individuals to participate in cancer screening. A previous study reported that a physician recommendation is the strongest predictor of undergoing cancer screening in patients with psychiatric disorders.^[24] However, most individuals with SPD reportedly do not use mental health services in Japan^[25]; therefore, strategies to increase cancer screening rates among them need to be developed. Further studies are needed to develop, disseminate, and implement interventions that encourage participation in cancer screening by people with SPD.

In the present study, we identified other demographic factors independently associated with participation in cancer screening, whether SPD was present or absent, in multivariate logistic analysis (Table 2). In previous Japanese studies using CSLC data before 2010,^[26,27] people who were divorced, those covered by the National Health Insurance, and people with lower household incomes were also found to be less likely to participate in colorectal, gastric, and lung cancer screening. These findings were confirmed in the present study, using recent datasets from the CSLC including multiple survey years. Although not the primary purpose of the present study, this result suggested that gaps in the cancer screening rate owing to socioeconomic disadvantages may remain unchanged. Further efforts to eliminate these gaps are needed.

The present study had several limitations that should be considered. First, SPD status was not assessed by a structured interview. The CSLC does not identify psychiatric diagnoses in respondents; thus, we could not determine what type(s) of mental disorders most strongly affect participation in cancer screening. Second, the self-reported data may have overestimated cancer screening rates. Third, there was a time gap between answering the K6 and participating in cancer screenings within 1 year. Fourth, there may have been a selection bias. Individuals with more severe psychological conditions may have declined to participate in the CSLC or to complete the K6 questionnaire or undertake cancer screening, which might have led to underestimation of the effect of SPD on participation in cancer screening. Fifth, in the present study, we did not investigate factors behind the changes in the gaps in cancer screening rates. Thus, we did not determine the factors that widen or narrow the gaps, which could have provided clues to developing effective interventions. Finally, it is unclear whether disparities in cancer screening rates among people with SPD contribute to disparities in their mortality rates.

In conclusion, between 2007 and 2016, cancer screening rates increased similarly over time among people with and without SPD. However, the gaps in cancer screening rates between people with and without SPD remained unresolved. These findings suggest that recent public activities to promote cancer screening may have been effective in increasing cancer screening rates in both groups to a similar certain extent, but are so far insufficient to eliminate the gaps.

Funding

This work was supported by the Japan Society for the Promotion of Science (JSPS) Grant-in-Aid for Scientific Research (KAKENHI) JP17K09112.

Conflict of Interest Disclosures

MF has received lecture fees from Mochida, Eli Lilly, and Sumitomo Dainippon and personal fees from Iyaku (Medicine and Drug) Journal and Igaku-Shoin outside the submitted work. KK has received lecture fees from Mochida outside the submitted work. NY has received grants from Otsuka, Astellas, MSD, Pfizer, and Takeda outside the submitted work. He has also received lecture fees from Otsuka, Astellas, MSD, Pfizer, Meiji, Janssen, Hisamitsu, Sumitomo Dainippon, Mochida, Tsumura, Takeda, Taiho, and UCB Japan outside the submitted work. MI has received grants from Novartis outside the submitted work. He has also received lecture fees from Meiji, Mochida, Takeda, Novartis, Yoshitomi, Pfizer, Eisai, Otsuka, MSD, and Sumitomo Dainippon and personal fees from Technomics. The institution of Masatoshi Inagaki has received grant or research support from Eisai, Astellas, Pfizer, Daiichi-Sankyo, Takeda, and MSD outside the submitted work. The authors declare that they have no financial conflict of interest with regard to the content of this report.

Authors’ contribution

MF: Study concept and design, data interpretation, literature review, data analysis, and writing of the initial draft; YH: Study concept and design, data interpretation, and writing of the initial draft. MI and NN: Study concept and design, data interpretation, and writing of the manuscript. YY, RW, TE, MF, KK, YU, TN, and NY: Study concept, data interpretation, and writing of the manuscript.