Estimates for cervical abnormalities in Vanuatu
Ms Nina Fotinatos, 28 Ryan Street, Brown Hill, Ballarat, Victoria 3350. Fax: (03) 5320 1190; e-mail: firstname.lastname@example.org
Objective: To use the Pap smear to establish a recent prevalence of cervical abnormalities within a select population in Vanuatu, a developing country.
Methods: Cervical smears (n=907) were collected from Ni-Vanuatu women from both urban and rural islands within Vanuatu between August 2001 and September 2005.
Results: The prevalence of low-grade epithelial abnormalities for the total population was 2.9% and the prevalence of the high-grade epithelial abnormalities/cancer was 2.0%. There was a significant difference (p<0.05) in prevalence of high-grade epithelial abnormalities/cancer between the urban and rural populations sampled, with a higher prevalence in the urban population.
Conclusions: The prevalence of pre-cancer and cancer in Vanuatu is high compared with Victorian (Australian) statistics yet comparable with other developing countries with no cervical screening programs available.
Implications: This study will hopefully assist in future planning of women's health programs and relevant preventive strategies to combat cervical cancer in Vanuatu.
Cervical cancer (CC) is the second most common female cancer type globally after breast cancer and the most common cancer in some developing countries.1 Nearly half a million cases of CC are diagnosed annually and more than 250,000 women die each year from this disease.1 Cervical cancer is a malignant growth and division of cells at the cervix or opening of the womb. Eighty-three per cent of all new cases of CC occur in developing countries, where there is a substantial lack of screening and treatment facilities.1–4 According to a literature analysis performed by Yabroff et al. (2005), three factors contributing to soaring CC mortality statistics include high human papilloma virus (HPV) prevalence, lack of or infrequent screening, and advanced disease at diagnosis with under-use of recommended treatment.5 Developing countries have the highest incidences of CC in the world, including countries in Central and South America, the Caribbean, sub-Saharan Africa and parts of Oceania and Asia.1,3,6-10
Since the introduction of the Pap smear in the 1950s conventional CC screening has been very successful in finding early abnormal changes that can be dealt with using minimal treatment techniques.3,10–15 Access to effective cervical screening programs has been shown to directly decrease incidence rates globally. During the past five years, only 5% of women in developing countries have had a Pap smear during opportunistic screening compared with 40-50% of women in developed countries.2,4,10
Vanuatu consists of a chain of approximately 80 islands in the South Pacific Ocean and is classified as a developing country, relying heavily on international assistance in areas of health, education and basic infrastructure.16 The majority of the population is involved with subsistence or small-scale agriculture production.17 Sixty per cent of the population is under 25 years of age and there are limited employment opportunities. The unemployment rate is high (78.21%) and the majority of unpaid subsistence farmers are women.18 Literacy of girls aged between 15 and 24 is an ongoing problem, with low attendance in secondary education.18 Basic sanitation is available to half the population, primarily in the urban areas. Access to safe drinking water is also available to only 38.5% of the population, again mainly in the urban centres.18
There is poor data supporting an accurate incidence and prevalence of CC in Vanuatu.19,20 Stewart (1996) quotes CC mortality incidence of 5-7 per 100,000 women between 1988/89, from an annual Department of Health Report (1987). During this period, an estimated prevalence of low-grade epithelial abnormality (LGEA) was 2.6% and high-grade epithelial abnormality (HGEA) was 0.8%. A current estimation would further compare the situation in Vanuatu with that of developing countries. With current estimations, the Ni-Vanuatu health officials are better equipped to understand the seriousness of the disease and any changing patterns in their population over the past 10 years. Strategies could include the development of a conventional or alternative cervical screening program or educational programs targeting possible vaccine use. Prevalence estimations between two islands as performed in this study would highlight specific needs and problems not addressed in earlier studies.
A total of 907 Ni-Vanuatu women participated in the study between 2001-05. With the assistance of the Australian Aid Obstetricians/Gynaecologists, Pap smears were collected from the urban region of Efate and the rural island of Ambae. The participants were aged between 18-65 years. The first site was located on the island of Efate (Port Vila) specifically at Vila Central Hospital (VCH), incorporating the Women's Health Unit (antenatal care). The women attending VCH paid a standard $2 hospital fee to meet with the gynaecologist on duty for routine or high-risk assessments related to pregnancy. It was after this assessment that a Pap smear would be offered, including a consent and information form prior to any sample collection.
The second site was on the island of Ambae, a smaller island in the northern part of Vanuatu incorporating an existing World Health Organization (WHO) health initiative. The village women presenting on the island of Ambae were invited by WHO to participate in a sexually transmitted disease workshop.
International couriers handled all the domestic and international freight to and from Vanuatu. Both recruitment sites were sent Pap smear kits containing frosted glass slides, Endocervex™ brushes, Ayre spatulas, pencils, information forms in English and Bislama (most commonly spoken local language), request forms, cytological fixative spray and plastic slide holders. Batches of Pap smears were returned to the cytology department at St John of God Pathology, Victoria, for examination, including staining using the Papanicolaou method.
Pap smears were screened and reported by two Australian Society of Cytology qualified cytologists. A pathologist reported all abnormal Pap smears. Reports were entered into a Microsoft Access Database called the Vanuatu CC Registry. The results were assigned into the following categories: unsatisfactory, negative for malignancy, possible and definite LGEA, inconclusive and HGEA incorporating cancer. Flora types were also commented on as part of the Pap smear screening process.
HPV effects can be commonly seen within epithelial abnormal cell changes in Pap smear known as koilocytes, and regularly co-exist with pre-existing LGEA and HGEA lesions. A molecular biological technique, including polymerase chain reactions (PCR), may also be employed. However, this was not used because of cost and availability in this study.
The data were analysed with the online Survey Star statistical software package utilising a z-test for two populations to determine any significant difference, where p<0.05. This statistical test was appropriate for the data collected during the study and was limited to comparisons of proportions. Results are either reported as significantly different (higher or lower) where p<0.05 or not significantly different, where p>0.05.
Table 1 refers to the basic demographics for the sample population. Almost one-third of the women sampled were aged between 26-30 years. The Efate women reported fewer children (2.7) compared with women in Ambae (4.3). The average parity for the total sample population was 3.3 children. Flora including Trichomonas vaginalis was detected in Pap smears during the screening process. The prevalence of Trichomonas vaginalis in the total study population was 25.3%, with a significant higher prevalence in the urban population (43.3% vs. 14.3%). There was no significant difference between the prevalence of women attending formal education between Efate and Ambae. However, there were significantly more women in Efate continuing to secondary or tertiary education compared with Ambae.
Table 1. Basic demographics.
Table 2 refers to the squamous cell categories in the total sample. There were significantly more cases of HPV present in Efate (2.7%) compared with Ambae (0.6%).
Table 2. Squamous cell category results for the total sampled population and the individual islands of Efate and Ambae compared with VCCR averages.
|LGEA|| || || || |
The aim of this study was to determine the prevalence of cervical abnormalities in the South Pacific island of Vanuatu. Using Australian Aid doctors and local health care professionals, 907 conventional Pap smears were obtained from two geographical separate areas of the country. Current statistics are required to support a preventive strategy for CC.
Due to the present level of understanding about Pap smears and minimal preventive health initiatives targeting women, few women would understand the reason for returning or undergoing further testing for CC. Therefore, the greatest limitation in this study is the lack of histological or colposcopic follow-up on abnormal Pap smears results, which was attempted with little success. It is widely accepted that the gold standard for supporting a cytological Pap smear diagnosis is histology; however in Vanuatu, as in most developing countries, histological facilities are unavailable. In some countries, histological testing is possible through the generosity of developed country laboratories, but this is extremely limited. The cost of histology relative to cytology and that it requires a specialist pathologist would make this science a poor use of already scarce health funds.
Unsatisfactory Pap smears
There are many causes for Pap smears to be deemed unsatisfactory. Insufficient cellular material was the most likely cause for an unsatisfactory report (51.6%) in this study. Low cellular material is commonly associated with the Pap smear sampling technique (not applying enough pressure) and can be affected by hormonal changes to the cervix, in particular pregnancy.22 The collectors in Efate were gynaecologists or gynaecology registrars with more experience at performing Pap smears compared with the nurses used in Ambae, who were hired from the Vanuatu Family Health Association. There were a significant lower number of unsatisfactory Pap smears taken in Efate compared with Ambae. Many studies discuss the experience and specific training of the Pap smear collector as a contributing factor to unsatisfactory rates where ongoing practice and feedback of the sample satisfactory rate assists in improving the Pap smear technique.23 There was a three-fold increase in the unsatisfactory smears in the Ambae population compared with the Efate population, even though 75% of the Efate samples were from pregnant women whose hormonal changes may contribute to unsatisfactory smears.
The second likely cause of an unsatisfactory result was due to a marked inflammatory response obscuring cellular detail (45.2%). Organisms such as Trichomonas vaginalis cause inflammatory responses in Pap smears, including grotty backgrounds. The prevalence of Trichomonas vaginalis in this study population was 25%, which is comparable with statistics from similar studies of 17% and 27.5%.21
Other causes for an unsatisfactory report include poor preservation (16.1%) and excessive blood obscuring cellular detail (4.8%). The humidity in Vanuatu possibly contributed to poor preservation, as any delay in applying fixative would markedly affect the quality of the smears.21 Hairsprays containing carbowax were used as a substitute fixative when a cytology fixative was unavailable, but this is a poor alternative and invariably results in poor preservation.
Even though these issues contributed to a high unsatisfactory rate, the study had significantly lower unsatisfactory results compared with Stewart (1996) in a similar study population.21 In other developing countries, studies demonstrate varying rates of unsatisfactory smears.14,24 There was a significant higher number of unsatisfactory Pap smears in this study compared with the Victorian Cervical Cytology Registry (VCCR).25 This was an expected finding as Australia has a national cervical screening program with high standards of training of the smear collectors.
The negative results between the two subpopulations in this study were significantly different. This difference is influenced directly by the unsatisfactory results of the two subpopulations. If the unsatisfactory results were excluded from the analysis, the negative results between the two subpopulations would not be significantly different.
The study had a significantly lower prevalence of negative results compared with Stewart (1996), despite sampling similar populations. The high prevalence of unsatisfactory results (15.3%) presented by Stewart (1996) influenced its overall prevalence of negative results (80.3%).21 Thus, its negative results appear significantly lower than those of the present study (87.5%).
There was no significant difference between the prevalence of negative results from this study and that presented by Naud et al. (2001), who sampled 100 women in Brazil.24 Similarities between the two developing study populations include the age range and the continuity of the reporting system. Naud et al. (2001) had the added advantage of colposcopy and biopsy for a more thorough follow-up of patients, which was not available for the present study.24
This study had significantly lower prevalence of negative results compared with Wasti et al. (2004) in Pakistan.14 One explanation for this difference is the influence of religious beliefs followed by Islamic followers particularly concerning sexual behaviour. Religious beliefs in Islam provide strict guidance in areas of sexual behaviour by influencing factors such as first age of intercourse, marital relationships, extramartial affairs, polygamy and the numbers of sexual partners a woman has during her lifetime. Approximately 90% of the population in Vanuatu are of Christian faith compared with the predominately Islamic community in Pakistan. The Christian faith does not share the exact teachings and regulations as the Islamic followers or the strictness concerning these teachings. Both these studies had significant differences between the unsatisfactory categories ultimately contributing to these findings.
This study had a significantly lower prevalence of negatives compared with VCCR statistics.25 If the unsatisfactory results were not included in our total Pap smear comparison, there would not be a significant difference between our negative results and those of the VCCR (91.6%), which suggests that the unsatisfactory smears substantially explain these findings.25
Low-grade epithelial abnormality/human papilloma virus
There was no significant difference between the prevalence of LGEA in Efate (3.4%) compared with Ambae (2.0%). The current sample size may influence this finding. LGEA is predominately a young woman's disease (21-30), which may contribute to this finding as the mean age of participants in Efate was lower than the mean age of participants in Ambae.4
Current literature states that HPV is a necessary cause of CC, however not solely sufficient to cause this disease. Although not significantly different, this biological connection may explain the higher prevalence of LGEA lesions in Efate compared with Ambae given the significant difference in the rates of HPV between the two populations. It is well documented that a large portion of HPV infections are transient and only a minor percentage of infections persist and progress to more advanced stages of cervical dysplasia.27 Low-risk HPV subtypes (type 6 or 11) are strongly associated with 90% of genital warts and possible LGEA lesions.28 It is hypothesised that other exogenous (oral contraceptives, smoking, diet) and endogenous (genetic factors, host immune response) factors influence the oncogenic activity linked to this virus.26,29
The prevalence of LGEA lesions in this study was not significantly different to the results presented by other studies in developing countries in South America and Vanuatu.21,23 This study had a significantly higher prevalence of LGEA lesions compared with a Pap smear study of Islamic women in Pakistan, where religion and lifestyle have been suggested to directly decrease risk factors and possibly the prevalence of CC.14 There is an associated risk with CC and the number of sexual partners a woman has, however this is strongly discouraged in the religious beliefs of Islam. Hence, it may explain the significant decrease in LGEA abnormality prevalence in the study by Wasti (2004) compared with our study population.14
Compared with the statistics from the VCCR, the study had a significantly lower prevalence of LGEA. However, the LGEA rate is affected by the unsatisfactory rate. A hypothetical assessment with the unsatisfactory samples removed showed no significant difference between these two groups. As the results stand, this is an unexpected finding for a developing country versus a developed country such as Australia. In a population with no national screening program, untreated LGEA lesions may naturally progress to HGEA lesions, thus giving a falsely low prevalence of LGEA.
High-grade epithelial abnormality/cancer
The Efate population had a significantly higher prevalence of HGEA/cancer lesions compared with the Ambae population. Possible high-risk HPV subtypes (type 16 and 18) have a strong association with HGEA/cancer abnormalities and thus may contribute to the HGEA/cancer abnormalities found in the Efate population.15,29,30
The HGEA prevalence is not significantly different compared with a similar opportunistic screening program performed by Naud et al. (2001) of 100 women in Porto Alegro, Brazil.24 However, Mauad (2002) also performed a study in Brazil (n=1,044) where there were significantly fewer cases of HGEA found compared with our study.31
There is a significantly higher prevalence of HGEA in this study compared with Stewart (1996).21 Although these studies sampled similar populations, Stewart (1996) did not sample any women from a rural setting and had a very high proportion of unsatisfactory results, which may have affected the estimated HGEA prevalence.21 The time between the two studies suggests that LGEA lesions 10 years ago might now be presenting as HGEA lesions.
The overall study had a significant higher prevalence of HGEA compared with the VCCR CC statistics.25 This was an expected finding as without any current organised national cervical screening program in Vanuatu, abnormalities have not been detected, treated and eliminated from the population. Vanuatu lacks human and financial resources in this health area, making it impossible to sustain a conventional screening program such as that available in Australia.
The results of this study highlight the progression of CC, particularly HGEA and cancer abnormalities, over the past 10 years in Ni-Vanuatu women. This shift is comparable with statistics from other developing countries that have some of the highest incidences in the world. This preventable disease affects not only women in their prime, but potentially also affects the family unit including upbringing of many young children. This data raises the question of further investigations into alternative techniques such as HPV vaccines or visual inspection with acetic acid used in similar settings in developing countries.
Any long-term program initiated must include a compulsory element of health educational strategies to inform Ni-Vanuatu women about the risk factors of the disease, signs and symptoms and possible treatment options.32 At present, there is one facility in Port Vila (Vanuatu Family Health Association) performing Pap smears for a reasonable cost to Ni-Vanuatu women. However, this is not regularly accessed due to the lack of knowledge in this area of women's health. These Pap smears taken in Vanuatu are referred overseas as there are no laboratories established for cervical screening. Private facilities are predominantly used by non-native women, who have been educated about the risks associated with CC and the screening methods prior to residing in Vanuatu.33
The data show that the urban population was more at risk in developing HGEA/cancer than the rural population. Factors such as HPV prevalence and the changing population in an urban environment, such as tourists, differ from isolated rural populations.
Although the rural population had a lower prevalence of HGEA/cancer, it still experienced hardship in areas of transportation and health care access because of the natural geographical dislocation of the Vanuatu islands. Future health strategies should also take into consideration these factors and possibly implement regular mobile medical units with ongoing educational sessions for women in isolated regions. 34–36