Equity in Cervical Cancer Screening in Iran: A Cross-Sectional Analysis of the 2016 Nationwide Steps Survey


 Background

One of the most important concerns in every healthcare system is the achievement of equity in health. Studies have shown great socioeconomic and spatial disparity in cervical cancer incidence and mortality. This study aims to investigate equity in cervical cancer screening in Iran.
Methods

A cross-sectional study was conducted using data from the National Non-Communicable Risk Factors Survey in 2016 (STEPs 2016). Data on cervical cancer screening in addition to demographic and socioeconomic factors from 15975 women aged 18 and above were analyzed. Chi-Square test and logistic regression were used to assess the association of demographic and socioeconomic factors with cervical cancer screening. Equity in screening was assessed by concentration index, and GIS analysis was used to show the provincial indices on the map of Iran.
Results

Overall, 41·6% of surveyed women in Iran had undergone cervical cancer screening in their lifetime. Participation rate in cervical cancer screening programs varied between provinces; ranging from 7.6% in Sistan and Baluchestan to 61.2% in Isfahan. Being single, living in rural areas, young age, having no education, and being uninsured were associated with lower participation. Concentration index showed pro-poor inequity for the country and across all provinces, indicating that cervical cancer screening services are less utilised by high income groups. GIS analysis demonstrated spatial disparity across provinces of Iran.
Conclusions

Equity and participation in cervical cancer screening in Iran requires improvement. For this to be achieved, new policies shall have a stronger emphasis on the lowest and the highest socioeconomic population groups, while current strategies have mostly affected the people with middle socioeconomic status.


Abstract Background
One of the most important concerns in every healthcare system is the achievement of equity in health. Studies have shown great socioeconomic and spatial disparity in cervical cancer incidence and mortality. This study aims to investigate equity in cervical cancer screening in Iran.

Methods
A cross-sectional study was conducted using data from the National Non-Communicable Risk Factors Survey in 2016 (STEPs 2016). Data on cervical cancer screening in addition to demographic and socioeconomic factors from 15975 women aged 18 and above were analyzed. Chi-Square test and logistic regression were used to assess the association of demographic and socioeconomic factors with cervical cancer screening. Equity in screening was assessed by concentration index, and GIS analysis was used to show the provincial indices on the map of Iran.

Results
Overall, 41·6% of surveyed women in Iran had undergone cervical cancer screening in their lifetime.
Participation rate in cervical cancer screening programs varied between provinces; ranging from 7.6% in Sistan and Baluchestan to 61.2% in Isfahan. Being single, living in rural areas, young age, having no education, and being uninsured were associated with lower participation. Concentration index showed pro-poor inequity for the country and across all provinces, indicating that cervical cancer screening services are less utilised by high income groups. GIS analysis demonstrated spatial disparity across provinces of Iran.

Conclusions
Equity and participation in cervical cancer screening in Iran requires improvement. For this to be achieved, new policies shall have a stronger emphasis on the lowest and the highest socioeconomic population groups, while current strategies have mostly affected the people with middle socioeconomic status.

Background
Cervical cancer was the fourth most common cancer and the fourth leading cause of cancer deaths in women worldwide in 2018; accounting for 570,000 new cases and 311,000 deaths, equal to 6·6% and 7·5% of the global cancer burden respectively. 1 Rates are significantly higher in developing countries.
3 Over eighty percent of new cases and cervical cancer deaths occur in less developed regions of the world. 2 Iran is among the countries with low incidence rate for cervical cancer. The estimated average agestandardized incidence rate for the country is 2·5 per 100,000 women. 3 Despite the low rate, mortality to incidence ratio is relatively high at about 42%. The advanced clinical stage, at which most cervical cancers are identified, is responsible for the poor prognosis and higher mortality rates. 4 The standard secondary prevention method for cervical cancer is cervical cytology, which has been established for more than four decades in most countries. 5 However, studies have documented large discrepancies in the effect of cytology screening programs on cervical cancer rates reduction. 6,7 The extend of participation of at risk women in screening program has been considered as the primary contributor to these differences. 8 There are different barriers deterring women from participating in cervical cancer screening program.
These factors will eventually result in differences in cervical cancer screening uptake which tend to be seen as inequities in the service. 9 Despite the global growing concerns about health equity in recent decades, there is a lack of research examining the socioeconomic disparities in cervical cancer screening utilization in Iran. Hence, this study sought to assess the equity in cervical cancer screening in Iran according to the data obtained from the National Survey of The Risk Factors of Non-Communicable Diseases (STEPs 2016), and furthermore to identify sociodemographic factors that predict the participation in cervical cancer screening program.

Methods
A study was performed using data from the National Survey of The Risk Factors of Non-Communicable Diseases in Iran (STEPs 2016). This cross-sectional household survey was conducted at the national level through personal interviews at homes. To estimate the sample size, a proportional to size sampling plan was adopted using the least populated province of the country as the basis and calculating the sample size of others according to the population ratio of each province to the referenced one. Weighting methods were used for provinces with sample size of 800 or more. In order 4 to control for sampling effect and non-response error, each provincial sample size was increased by 10%. Ultimately 3105 clusters, each consisting of ten households, were selected from urban and rural areas by designing a systematic cluster random sampling frame for all provinces. Additional details on the methodology of the survey is available elsewhere. 10 Totally 31,050 participants aged 18 or older were interviewed nationwide. Among the interviewees 15,970 participants were women. All the women were asked the following pap smear screening question: "Have you ever been screened for cervical cancer?" or "Have you ever had a Pap test?".

Variables
In order to assess the determinants of participation in cervical cancer screening program, the outcome variable was defined as having ever undergone cervical cancer screening (yes/no). The independent variables selected for the analysis included age, marital status, education level, employment status, residing area, health insurance coverage. For easier interpretation, all variables were categorized. Age was categorized into ten-year age groups starting from the age of 18 years, and education level was classified by number of years attained. Additionally, employment status was defined as either employed, unemployed, retired, or student. Residing area was categorized as urban or rural, and insurance coverage as having been covered or not.
In order to measure equality, a single living standard variable was produced through principal component analysis (PCA) statistical method and by using data from household assets. Asset variables included home, car, kitchen, bathroom, phone, TV, air cooler, central heating system, refrigerator, freezer, oven, vacuum cleaner, washing machine, dishwasher, personal computer, mobile phone, internet access, and having access to water and gas pipelines. To check if the data is suitable for data reduction, the following two tests, Kaiser-Mayer-Olkin (KMO) Measure of Sampling Adequacy and Bartlett's test of sphericity, were carried out prior to PCA. The value of KMO obtained was 0·787 and the result of Bartlett's test of sphericity was significant, therefore PCA was appropriate. 11 In the analysis performed, seven components had eigenvalues greater than 1 and explained 64% of the variation in the data. A socioeconomic status variable was created by retaining the principal component with the largest eigenvalue. The constructed variable was then divided into five quintiles, where the first and the last quintile represented the poorest and the richest socioeconomic levels respectively.

Statistical Analysis
Descriptive statistics (absolute and relative frequencies) were used to describe the distribution of women with regard to cervical cancer screening practice. To examine the association between sociodemographic characteristics and the cervical cancer screening adherence, the Chi-Squared test was performed. Logistic regression analysis was also undertaken for calculating adjusted sociodemographic disparities. The estimates were analysed by using SPSS version 25, assuming a significance level of α < 0·05.
Concentration curve and concentration index (CI) were used to assess the inequality in cervical cancer screening participation according to the socioeconomic status in Iran. The concentration curve plots the cumulative percentage of health outcome (y-axis), against the cumulative percentage of the individuals ranked by socioeconomic status (x-axis). The concentration index is measured based on the concentration curve and takes values between − 1 and + 1. In the case of perfect equality, the curve is a 45° line and the index equals to zero. If the health variable is distributed disproportionately higher among the individuals with higher socioeconomic level, the curve lies below the line of equality and the index takes positive values. The curve lies above the line of equality when the health variable is unequally distributed in favour of individuals in lower socioeconomic level. In this case the index is negative. 12 In this study for the binary health outcome, which was having ever undergone cervical cancer screening or not, the Wagstaff correction was needed in order for concentration index to be measured between − 1 and + 1. 13 Therefore, inequality analyses were performed considering the Wagstaff correction by using STATA version 15.
The concentration index was also calculated for the provinces of the country. Moreover to provide more insight of the spatial distribution, Geographic Information System analysis (GIS) was used to visualize the results on the map of Iran.

Results
Data from 15975 participants were analyzed. The characteristics of the study participants are shown 6 in table 1. Most women in our sample were aged 25-34, married, and had elementary education (1-6 years). The majority were residing in urban areas and unemployed. Over ninety percent had basic primary health insurance, yet less than a third were covered under the complementary insurance plan.
Overall, 41·6% of women reported ever had a cervical cancer screening. The highest proportion was found in women aged 35-44 years. Women who were married, had 6-12 years of schooling, lived in urban areas, were unemployed, and had insurance coverage, showed greater participation in cervical cancer screening program. All sociodemographic variables were significantly associated with cervical cancer screening in the bivariate analysis, except for the age group. (Table 1) Table 2 shows participation rate for each province as a whole and by urban and rural populations. As shown, highest rates were obtained in the provinces of Isfahan, Fars and Kohgiluyeh and Boyer-Ahmad. In contrast, Sistan and Baluchestan, Hormozgan and Kerman had lowest participation rates.
Urban-rural disparity in participation rate was greatest in Khorasan-Razavi and Ilam, while being smallest for Fars and Kohgiluyeh and Boyer-Ahmad.
The results of the multivariate analysis of the factors associated with participation in cervical cancer screening are presented in table 3. Positive predictors for having ever had a cervical cancer screening were age 35-44 years, married, 6-12 years of schooling, basic and complementary health insurance.
Living in rural areas was negatively associated with cervical cancer screening. Moreover, women who were employed or university student, were less likely to participate in cervical cancer screening program.
Analysis of the rate of cervical cancer participation in women across the socioeconomic levels revealed that participation rate was greater among women in lower socioeconomic levels as compared to that among those in higher levels. The concentration index was obtained as -0·22, indicating that the health service was concentrated in lower socioeconomic levels of the society. The corresponding concentration curve is shown in figure 1.  which has been in place for more than three decades, has failed to reach the majority of at-risk women in the population.
Age was significantly associated with cervical cancer screening participation. The proportion screened increased with age until 35-44-year age group, but then declined in each subsequent age group afterward. This was aligned with other studies in Iran and elsewhere. 15, 20-22 Women in younger age 8 probably tend to be more informed about cervical cancer and are more likely to visit healthcare providers for gynecological complaints.
As expected, married women were most likely to have undergone cervical screening, which was in line with previous studies. 15,18 Married women tend to attend healthcare facilities more often for maternal and child health care.
With regards to education level, higher education was associated with higher participation until about 12 years of schooling. This finding is consistent with other study results conducted in Iran, but not with those of other countries. 15,18,22,23 The finding suggests that university education in the country has no impact on raising women's awareness or changing their attitude towards cervical cancer screening.
The present study determined that employment or being a university student were negative In parallel with other studies, women with health insurance coverage were more likely to have undergone cervical screening. 15,25,26 Despite of the considerable success of the Iranian primary health care approach in delivering health services to remote and rural areas of the country in the past three decades, women who resided in rural areas had lower probability of being ever screened. 27 This may be partly due to the barriers such as embarrassment and lack of privacy. Since the screening service is provided by a community health care worker who lives in the same neighborhood.
This study also revealed evidence of disparities in cervical cancer screening participation across socioeconomic groups in Iran. Contradicting with findings in most countries, women with higher socioeconomic status had shown lower participation in cervical cancer screening program. 14,17 However, pro-poor inequality was shown in outpatient healthcare utilization in multiple studies in Iran. 28,29 Perhaps this finding could be explained by the nature of the health care system in Iran. In the absence of family physician program and a referral system, women in higher socioeconomic levels often visit specialists directly and bypass primary health care facilities. 30 Regional variations in equity in cervical cancer screening participation could be understood by portioning the concentration index analysis across provinces. Provinces with higher socioeconomic status, less cultural conservatism, and better social health represent greater equity in cervical cancer participation.
The major strength of the study is the use of a large nationally representative sample of rural and urban Iranian women, containing different socioeconomic levels, that allowed more confident inferences about the population. However, there are some potential limitations as well. First, the cross-sectional design of the study has limited the ability to draw conclusions about causal relationships. Second, the data on the history of having ever undergone cervical cytology were selfreported, and therefore may be susceptible to recall and social desirability biases. Lastly, by evaluating secondary data sources in this study, the assessment of all factors associated with cervical cancer screening participation was not possible.

Conclusion
Considering the findings of this study, it can be concluded that participation rate for cervical cancer screening is not optimal in Iran and could be improved. Moreover, with regards to the distribution of cervical cancer screening practice, social and geographical disparities have been determined which indicate the need for further research and more comprehensive strategies in order to reduce them.

Consent for publication
Not applicable.

Availability of data and materials
The data that support the findings of this study are available from National Institute of Health Research (NIHR) but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of National Institute of Health Research (NIHR).

Conflict of Interest
The authors declare that they have no competing interests associated with the material presented in this paper.

Role of the Funding Source
This project has been supported for data cleaning, analysis, and interpretation by the Research Deputy of the School of Medicine at Shahid Beheshti University of Medical Sciences in Tehran, Iran (Project no. 18109). There was no financial support for authorship and publication.

Contribution of the authors
RA was the main researcher and involved in study design, literature search, data analysis, data interpretation, article drafting and finalizing the manuscript. AAK was involved in data cleaning, study design, data interpretation and article drafting. NJ and ARA were involved in study design, data interpretation, data analysis and article drafting. MRS was the head of team and involved in study design, literature search, data analysis, data interpretation, article drafting and finalizing the manuscript and is the corresponding author. All authors read and approved the final manuscript. percentage of women in each category who reported undergoing cervical cancer screening; P-value, obtained from cross-tabulation between each category and cervical cancer screening.    Provincial Concentration Index for Cervical Cancer Screening Participation in Iran.