The global burden of cancer continues to rise with about 14.1 million new cancer cases, 8.2 million cancer deaths, and about 32.6 million people living with the disease worldwide (within 5 years of diagnosis) in 2012 [
Although it is already a major public health concern, the burden of prostate cancer in SSA is expected to grow mainly due to growth and aging of population, changing diets, lifestyles, and socioeconomic conditions [
While prostate cancer is the most common kind of cancer among men of African descent [
In Namibia, prostate cancer accounts for 44.8 per 100,000 of all cancers among men. It also accounts for 21.2 per 100,000 of overall cancer incidence and 21.5 per 100,000 of all cancer mortalities among adult men. Further, the country has a 5-year prevalence rate of 28 per 100,000 [
Factors such as inadequate public health infrastructure and other health concerns such as HIV/AIDS and malaria compete for scarce health resources and likely undermine the provision of prostate cancer services. These challenges have similarly been observed in cervical cancer screening in Namibia [
Namibia’s health insurance (medical aid) scheme relies on government and private not-for-profit organisations to manage health financing. The private not-for-profit organisations mostly provide “high-option” products and extensive coverage for inpatient and outpatient services to their voluntarily registered members under two main schemes referred to as open and closed schemes [
In a comparative study of Canada and United States [
This study used the 2013 Namibia Demographic and Health Survey (NDHS), a nationally representative dataset collected jointly by the National Statistical bureau and Ministry of Health of Namibia and MEASURE DHS program in Calverton, Maryland, USA. The NDHS is administered face to face to men aged between 15 and 64 years and collected periodically to provide data on basic national demographic and health indicators to guide policy makers, planners, and researchers. It is one of the few national surveys in SSA which has recently introduced a set of indicators on prostate cancer screening in order to assess the prevalence and risk factors in the general population. The current study focuses on a subsample of 1,244 men aged 40 and above.
The outcome variable of this study, prostate cancer screening, is a binary dependent variable measured with the question “have you ever been examined for prostate cancer?”, coded “0” for no and “1” for yes. The main explanatory variable of the study—health insurance coverage—was constructed from the question “are you covered by health insurance?”, coded “0” for not covered and “1” for covered. To capture the role of capacity for health literacy, the study also included a variable on education “0” for no education, “1” for primary, “2” for secondary, and “3” for higher and whether men discussed family planning issues with a health worker in the last 12 months “0” for no and “1” for yes. The role of health literacy was further explored by the variable tapping into exposure to media in which men were asked as to whether they listen to radio coded “0” for not at all, “1” for often, and “2” for very often or watch television coded “0” for not at all, “1” for often, and “2” for very often. This was important given the use of mass media in the dissemination of medical information in Namibia and other SSA countries. The analysis also examined the mediating effect of socioeconomic status using wealth quintiles. Wealth is a composite index created based on a household’s ownership of a number of consumer items which the NDHS deems to be poorest, poorer, middle, richer, and richest quintiles and recoded “0” for poorest and poorer; “1” for middle; and “2” for richer and richest. Demographic variables included in the analysis are age of respondents in 5-year categories, marital status coded “0” for single; “1” for married, and “2” for separated, and religion coded “0” for Catholics; “1” for Protestants; “2” for ELCIN (a type of Christian religion practiced in Namibia); and “3” for other religious groups. Locational factors controlled for include place of residence coded “0” for urban, “1” for rural and geographic region of residence coded “0” for Caprivi, “1” for Erongo, “2” for Hardap, “3” for Kara, “4” for Kavango “5” for Khomas, “6” for Kunene, “7” for Ohangwena, “8” for Omaheke, “9” for Omusati, “10” for Oshana, “11” for Oshikoto, and “12” for Otjozondjupa. The reference categories of all variables are coded “0.”
We used complementary log-log models instead of binary logit model to analyze our outcome variable given the highly uneven split of the outcomes in the dependent variable (see Table
Sample characteristics.
Variable | Frequency (%) |
---|---|
Ever tested for prostate cancer | |
No | 1,044 (83.92) |
Yes | 200 (16.08) |
Health insurance (ref: none) | |
No | 841 (67.60) |
Yes | 403 (32.40) |
Education (ref: none) | |
No formal education | 226 (18.17) |
Primary | 420 (33.76) |
Secondary | 481 (38.67) |
Higher | 117 (9.41) |
Discussed health issues with health worker in the last 12 months | |
No | 1,185 (95.26) |
Yes | 59 (4.74) |
Frequency of reading newspapers | |
Not at all | 1,000 (80.39) |
Often | 145 (11.66) |
Very often | 99 (7.96) |
Frequency of listening to radio | |
Not at all | 178 (14.31) |
Often | 274 (22.03) |
Very often | 792 (63.67) |
Age of respondent (mean) | 49 |
Marital status | |
Single | 239 (19.21) |
Married | 899 (72.27) |
Separated | 106 (8.52) |
Religion | |
Catholic | 298 (23.95) |
Protestants | 203 (16.32) |
ELCIN | 515 (41.40) |
Others | 228 (18.33) |
Region of residence | |
Caprivi | 58 (4.66) |
Erongo | 150 (12.06) |
Hardap | 127 (10.21) |
Karas | 126 (10.13) |
Kavango | 76 (6.11) |
Khomas | 99 (7.96) |
Kunene | 89 (7.15) |
Ohangwena | 48 (3.86) |
Omaheke | 126 (10.13) |
Omusati | 76 (6.11) |
Oshana | 61 (4.90) |
Oshikoto | 138 (11.09) |
Otjozondjupa | |
Place of residence | |
Urban | 615 (49.44) |
Rural | 629 (50.56) |
Wealth | |
Poorest | 172 (13.83) |
Poorer | 215 (17.28) |
Middle | 248 (19.94) |
Richer | 304 (24.44) |
Richest | 305 (24.52) |
Bivariate analysis of prostate cancer screening (complementary log-log).
Variable | OR (SE) |
---|---|
Health insurance (ref: none) | |
Yes | 6.77 (1.172) |
Education (ref: none) | |
Primary | 2.43 (.976) |
Secondary | 6.75 (2.556) |
Higher | 22.64 (9.119) |
Discussed health issues with health worker in the last month (ref: none) | |
Yes | 2.51 (.746) |
Listen to radio (ref: none) | |
Often | 0.38 (.125) |
Very often | 0.58 (.196) |
Watch television (ref: none) | |
Often | 1.45 (.435) |
Very often | 1.49 (.391) |
Age of respondent (ref: 40–44) | |
45–49 | 1.38 (.312) |
50–54 | 1.87 (.443) |
55–59 | 2.08 (.522) |
60–64 | 2.10 (.556) |
Marital status (ref: single) | |
Married | 3.85 (1.141) |
Separated | 2.83 (1.114) |
Religion (ref: Catholic) | |
Protestants | 1.71 (.455) |
ELCIN | 1.19 (.272) |
Others | 2.14 (.518) |
Region of residence (ref: Caprivi) | |
Erongo | 2.49 (1.331) |
Hardap | 2.08 (1.150) |
Karas | 3.00 (1.616) |
Kavango | 1.15 (.727) |
Khomas | 2.63 (1.466) |
Kunene | 0.56 (.381) |
Ohangwena | 2.68 (1.660) |
Omaheke | 1.46 (.833) |
Omusati | 0.66 (.451) |
Oshana | 1.23 (.789) |
Oshikoto | 1.08 (.694) |
Otjozondjupa | 2.44 (1.324) |
Place of residence (ref: urban) | |
Rural | 0.49 (.091) |
Wealth (ref: poorest) | |
Poorer | 1.16 (.581) |
Middle | 1.74 (.796) |
Richer | 4.03 (1.682) |
Richest | 13.61 (5.498) |
Standard errors are in parenthesis.
Factors associated with prostate cancer screening (complementary log-log).
Variable | Model (1) | Model (2) |
---|---|---|
Health insurance (ref: none) | ||
Yes | 4.11 (.787) |
2.95 (.620) |
Education (ref: none) | ||
Primary | 1.98 (.877) | 2.08 (.927) |
Secondary | 4.01 (1.568) |
3.32 (1.392) |
Higher | 8.13 (3.450) |
6.34 (2.915) |
Discussed health issues with health worker in the last month (ref: none) | ||
Yes | 1.54 (0.455) | 2.02 (.611) |
Listen to radio (ref: none) | ||
Often | 0.91 (.379) | 0.96 (.404) |
Very often | 1.10 (.473) | 0.99 (.440) |
Watch television |
||
Often | 1.12 (.338) | 1.39 (.445) |
Very often | 0.94 (.250) | 0.83 (.232) |
Age of respondent |
||
45–49 | 1.24 (.278) | |
50–54 | 2.05 (.504) | |
55–59 | 2.06 (.532) | |
60–64 | 3.30 (.990) | |
Marital status (ref: single) | ||
Married | 1.56 (.476) | |
Separated | 1.87 (.737) | |
Religion (ref: Catholic) | ||
Protestants | 1.02 (.275) | |
ElCIN | 1.18 (.271) | |
Others | 1.12 (.278) | |
Region of residence |
||
Erongo | 1.25 (.677) | |
Hardap | 1.33 (.750) | |
Karas | 1.71 (.920) | |
Kavango | 1.52 (.958) | |
Khomas | 1.18 (.661) | |
Kunene | 0.82 (.565) | |
Ohangwena | 5.10 (3.222) | |
Omaheke | 1.29 (.734) | |
Omusati | 0.47 (.320) | |
Oshana | 0.89 (.550) | |
Oshikoto | 1.03 (.653) | |
Otjozondjupa | 2.11 (1.159) | |
Place of residence (ref: urban) | ||
Rural | 1.51 (.327) | |
Wealth (ref: poorest) | ||
Poorer | 1.12 (.586) | |
Middle | 1.25 (.628) | |
Richer | 2.41 (1.186) | |
Richest | 4.95 (2.613) | |
Random effect | ||
Variance at the cluster level | 2.106 (.353) |
1.56 (.416) |
Constant | 0.021 (.009) |
0.003 (.003) |
|
||
Observations | 1,244 | 1,244 |
Standard errors are in parenthesis.
Our findings indicate that only 16% of men reported ever screening for prostate cancer in Namibia. About 32% of our sample reported having health insurance and only 5% of men reported ever discussing family planning issues with a health worker in the last 12 months before the survey. About 39% of men had a secondary education, 72% reported being married, with a mean age of 49, and about 41% of the sample were identified with the ElCIN religion. The distribution of men with regard to place of residence was near even between urban and rural areas, 49% of men residing in urban areas and approximately 51% in rural areas. About 50% of men were within the richest and richer wealth quintiles whilst about 31% were in the poorest and poorer categories. The middle wealth quintile accounted for about 20% of the sample.
The results of the bivariate logistic models are reported in Table
Two multivariate results are presented in Table
The second model controls for demographic and socioeconomic variables. We found that the association between health insurance coverage and prostate cancer testing attenuated after adjusting for socioeconomic and demographic variables but remained significant and robust. Other variables associated with screening included level of education, age of respondent, contact with health personnel, region of residence, and wealth category. Compared to men without formal education, men with secondary (OR = 3.32,
We examined the determinants of prostate cancer screening among men of 40 years and over, considered to be the age group at risk of prostate cancer. Our findings show that Namibian men with health insurance coverage, having access to information, having contact with health workers, and residing in richer and richest wealth quintiles, were more likely to screen for prostate cancer. The effect of health insurance on testing for prostate cancer remained robust even after controlling for access to information and socioeconomic and demographic factors, suggesting the disproportionate influence that having insurance coverage might have on an individual’s access to cancer screening. This particular finding is generally consistent with the literature on the effect of insurance coverage on health utilisation in different places [
Invariably, given the relative contribution of insurance coverage and wealth to prostate cancer screening in this context, it means that the poor face a dual burden of poverty and inequity in health access. Hence, the poor are more likely to be uninsured and are also more likely to face barriers to preventive information on prostate cancer screening. This may be due to access disparities among insured and uninsured individuals, often rooted in income inequalities [
An interesting finding of this study is the positive relationship between discussing health issues with a health worker and screening for prostate cancer. This suggests that the appropriate promotion of prostate cancer screening through health workers will be useful to encourage men to test especially in a context where reproductive health services have historically been directed at women. There is the need to push for more openness and awareness in order to promote dialogue between health professionals and men on relevant issues around prostate cancer and encourage them to screen for prostate cancer. Our results are consistent with those of other studies which have reported that individuals who make regular visits or are in regular contact with health worker(s) tend to be better informed about health issues, are familiar with medical settings, are more receptive to medical advice, and are more likely to undergo testing [
The progressive association between age and testing for prostate cancer may be a reflection of more positive behaviours to learn about risk factors and willingness to adopt preventive measures such as screening in order to seek treatment. This particular finding is generally consistent with other studies that have singled out age as one of the widest known risk factors for developing prostate cancer alongside ethnicity and race [
The positive association between wealth and testing for prostate cancer is noteworthy. The richer and richest categories were more likely to report testing for the disease, reemphasising the notion that it is mostly those who have the financial means to overcome barriers to health care services. This is similarly the case in the context of health insurance coverage where the richer and richest tend to have better access to prostate cancer testing. The relatively low likelihood of testing among the poor highlights the issue of socioeconomic inequalities to cancer screening and underscores the kinds of barriers that poor people face in terms of access to testing. Since testing is a gateway to treatment, the findings of this study also suggest potential socioeconomic disparities in morbidity and mortality from cancer in Namibia. Furthermore, even though prostate cancer screening is generally low in Namibia, the findings of this study suggest existence of wide geographical variations in terms of screening. For instance, residents in Ohangwena region were more likely to screen for prostate cancer, compared to Caprivi, one of Namibia’s poorest and underserved regions [
This study has some limitations. First, due to the cross-sectional nature of the dataset, we are unable to make causal linkages between prostate cancer screening and any of our independent variables. Also, due to the self-reported nature of the data, some biases may have been introduced into the data during data collection as men are more likely to provide socially satisfactory responses and the NDHS could not physically validate these responses. We do also acknowledge that GLOBOCAN data on mortality are projections and may overstate or understate the burden of prostate cancer in Namibia. Furthermore, even though prostate cancer is deserving public health attention, it should be noted that there still remain other more common noncancer causes of mortality and that knowing one’s prostate cancer status will not necessarily prevent death. Such considerations should be factored in when prioritizing public health policies in such limited resources settings that have to prioritize their objectives. To a large extent our results are generalizable to other resource-poor countries in Sub-Saharan Africa, even though one must not lose sight of the contextual influence of culture, norms, health behaviours, and the political support of Namibian government in prioritizing population based screening.
In conclusion, this paper has examined the determinants of prostate cancer among men aged 40–64 years in Namibia. The significant role played by health insurance coverage in influencing screening highlights the need for a national health insurance strategy that ensures equity in health access, especially screening for cancer. Currently, Namibia does not have a universal health insurance policy although discussions are currently underway to introduce a national scheme to reduce out-of-pocket health expenses and inequities in access to health services. It is hoped that this may impact positively on health care utilisation including prostate screening. We also urge that for such a scheme to be effective in increasing screening for prostate cancer, it has to be accompanied by a strong health promotion campaign to promote the public awareness about the disease. The study also points to the role played by regular contact with health workers in promoting testing for prostate cancer among men, underscoring the need for the government to reduce barriers that make it difficult for people to get in touch with health personnel or to have a regular doctor. It also suggests the need for a cultural shift that would promote more dialogue on men’s reproductive health issues in a context where women have traditionally been the subject of such debates. The current recommendations for prostate cancer screening are more appropriate for developed country contexts as they have the resources and technical expertise to handle the burden of prostate cancer. In a resource limited setting such as Namibia, outcomes are likely to be poorer and there is also the strong likelihood that many patients that would be correctly diagnosed may not receive treatment due to lack of appropriate resources. This is especially the case in Namibia where, due to a lack of previous testing, the national rollout of screening is likely to uncover many cases of advanced prostate cases which may be difficult to treat. As a public health consideration, the Namibian government should carefully consider likely benefits from the national screening program with respect to its capacity to provide appropriate care for those who test positive. Those making decisions about commitment of public health resources need to weigh the costs associated with prostate cancer against those of other public health interventions, such as HIV/AIDS, malaria, or even cervical cancer, whose diagnostics and interventions are relatively low cost. Cervical and breast cancer are also other important and competing public health problems in Namibia.
Screening for prostate cancer remains low in Namibia with only 16% of men reporting having ever tested. Men with health insurance and those who discuss their health issues with a professional were more likely to screen for prostate cancer; the findings suggest that expanding health insurance coverage together with prostate cancer screening education could improve the outcomes. The study contributes to the current field of knowledge of prostate cancer testing among resource-poor populations given the high risk of prostate cancer within these settings.
The sponsor has no role in the study design, data analysis and interpretation, writing the paper, and the decision to submit the paper for publication.
The authors have no conflict of interests to disclose.
The authors are grateful to MEASURE DHS for granting them access to use the Namibia DHS. Joseph Kangmennaang gratefully acknowledges funding support from the Canadian Queen Elizabeth II Diamond Jubilee Scholarships for his graduate education.