514 SOCIAL SUPPORT AND CERVICAL AND BREAST CANCER SCREENING PRACTICES AMONG NURSES

This cross-sectional epidemiological study was carried out at three public hospitals in Rio de Janeiro, RJ, Brazil. It aimed at analyzing the association between social support and cervical and breast cancer early detection practices among nurses. Data were collected through a multidimensional questionnaire focusing on socio-demographic and occupational variables, the performance of the breast self-exam, the Papanicolaou smear test, and social support (Medical Outcomes Study scale). Statistical associations were evaluated through the Chi-square test (p d” 0.05). Logistic regression tests were used for multivariate analysis. Higher levels of social support consistently increased the chances of individuals reporting adequate practices concerning breast self-examination and having Pap smear tests performed, regardless of socio-demographic or occupational variables. These results corroborate the hypotheses that social support has a positive effect on the regular practice of self-care.


INTRODUCTION
With more than 10 million new cases every year, cancer has become one of the most threatening diseases in the world (1) .Breast cancer is the most frequent type of cancer among women and probably the most feared due to its frequency and psychological effects related to the perception of sexuality and a person's self-image (2)   .Cervical cancer is the second most common type of cancer in Brazil (1) .
Due to the high rates of mortality and physical and emotional sequelae these two types of cancer cause, their prevention and early identification is of great importance in considerably improving the probability of cure (1) .Early detection requires, on the one hand, health services to provide screening exams and, on the other hand, the voluntary search for these services.However, several factors can influence adherence to breast and cervical cancer detection practices.Among these aspects, the ones related to socioeconomic conditions are highlighted such as: schooling, income, residence in rural or urban areas (4) .
Other factors are related to women's access to health programs and services that provide screening exams, access to regular health service or a professional of reference, and also women's perceptions of their own health conditions, history of cancer among acquaintances or family members and lifestyle (4) .In addition to these factors, several international studies (5)(6)(7)(8)(9) have stated that social support also promotes preventive health care.
Social support is related to resources supplied by the individual's social network when in need (10) and can be measured through the individual's perception of the degree to which interpersonal relationships correspond to certain functions, i.e. emotional, material and affective support (11) .The power of social networks theory is based on the assumption that the social network structure, by itself, is highly responsible for determining individual attitudes and behavior through access to resources and opportunities, and stimuli to behaviors (9)(10) .The central idea is that lifestyle, exposure to certain stressful events in life, chronic stressing experiences and psychological resources of individuals are generated in the context of the social structure in which people live (10) .
In the context of studies in Brazilian nursing, the importance of social support has been recently emphasized (11)(12)(13)(14) .However, few national studies evaluate the relation of social support to women's preventive practices.Acknowledging the importance of social support in different health contexts is essential both for planning nursing care and for developing studies that can be applied in such care (12) .
This Support Survey (MOS-SS), originally in English (15) , translated to and validated in Brazilian Portuguese (16)   .
It is a 19-item scale comprised of five functional dimensions of social support (15) : material (four questions -supply of practical resources and material to them in case of need: "none of the time", "a little of the time", "some of the time", "most of the time", "all of the time" (15-16)   .
Although the social support scale was originally developed to measure five dimensions, the factorial analysis showed, both in its original version (15)   and for the Brazilian (16) context, that the emotional dimension could not be discriminated from the informational dimension.Additionally, the results from the evaluation of the Brazilian version (16) indicated that the items of positive social interaction support could not be discriminated from affective support items, and constituted a single dimension (affective/positive social interaction).Hence, we opted, based on the study analyses, to use three dimensions of support: 1-"material", 2-"emotional/informational" and 3-"afective/positive social interaction (PSI)".To facilitate comparisons between these dimensions, we standardized them using the ratio of the obtained scores from the set of questions of each dimension and the maximum values of possible scores according to the number of questions.The result of this ratio was multiplied by 100 (11)(12) .Then, this score was divided into terciles, indicating that the higher the obtained score, the higher the level of social support.
The breast self-exam (BSE) was investigated through the question: How frequently do you selfexamine your breasts, aiming to check for lumps or any abnormality?Responses were classified into two categories (17) : 1 -practice -includes all those who perform BSE "every month" or "almost every month" and 2 -do not practice -includes all those who "rarely" or "never" perform the BSE.
The frequency participants have a Papanicolau smear done was investigated through the question When was the last time you had a preventive exam for cervical cancer done?Responses were again classified into two categories: 1 -inadequate practice -for those who answered "never have", "between two and three years", "between three and five years" and "for more than five years" and 2 -adequate practice -for those whose answer was "less than one year" or "between one and two years" (18) .Univariate and bivariate analyses were performed to verify the association between social support and screening practices evaluated with each of the studied sociodemographic co-variables (age, schooling, marital status, race/ethnicity, per capita income, children) and occupational co-variables (professional category at the hospital, another job, type of employment contract, years working in nursing and weekly work hours).In this stage, the Person's Chi-square test was used to verify if associations presented statistical significance at 10% (p<0.10).This criterion was used to select potential confounding variables.Those variables which were found to be associated both with the exposure and the outcome were included in the multivariate models through logistic regression.
The research project was approved by the Ethical Committees at the hospitals where the study was conducted and also by the CONEP in Brasilia (No. 10228) because it is partly supported by an international institution.

RESULTS
Of the studied nursing workers, 28.4% were between 46 and 55 years old, 43.3% were married, 59.6% had children, 53.9% had completed high school, 36.8% self-reported as being white and 31.8% had a per capita income between R$ 350 and R$ 699 (Table 1).showed that the youngest, with the highest level of schooling, highest income, married, without children and who self reported as being white obtained the highest scores of social support (these results are not presented in the tables).
In relation to female cancer screening practices, we observed that almost 83% had a Pap smear test done between one and two years.However, 3.7% had never been screened and 12.7% had it performed more than two years before.In regard to the BSE, 8.4% reported they had never performed it, 53.9% rarely or sometimes performed it and 36.6% performed it almost every month or every month (Table 3).In the multivariate analysis, adjusted by sociodemographic and occupational variables, considered confounding variables, the odds of having a Pap smear done was 58% higher (OR=1.58;CI=1.07-2.34) in the 3rd percentile (higher levels of support) when compared to the 1st percentile (lower levels of support) of social support.
Concerning the dimension of positive social interaction/affective social support, the odds of having a Pap smear done among those who showed intermediary levels of social support was 70% higher (OR=1.70;CI 95%=1.16-2.47)and for those with higher levels of social support (3 rd tercile), the odds was 84% higher (OR=1.84;CI 95%=1.24-2.71),as compared to workers classified in the group with the lowest positive social interaction/affective social support (first tercile).
It was not possible however to identify any association between material support and emotional support/information and the Pap smear test (Table 4).Analyses of association of support social levels with BSE, adjusted by the confounding variables, showed that the odds of performing the BSE was 54% higher (OR=1.54;CI95%=1.14-2.07) in the last tercile of global social support than in the first.
Additionally, those workers classified in the third tercile of emotional support/information and in the third tercile of the PSI/affective had higher odds of performing the BSE (43% and 80% higher, respectively), compared to the first tercile group, which confirms that there is a direct association between the scores of the global social support dimensions and BSE (Table 5).
Similar to the Pap smear test, material support was not associated to greater/lesser chances of performing the breast self-exam among the evaluated workers (Table 5).empowerment of health care as it facilitates access to information and transference of knowledge, and provides encouragement (9) .Other authors (10) also point to the importance of social influence in which the social environment would establish somewhat acceptable behavioral standards, while certain social ties would promote higher access to health services.

CONCLUSIONS
This study identified a considerable proportion of nursing workers with inadequate preventive practices in relation to cervical and breast cancer screening practices, showing that prevention opportunities are being lost in the context of the studied hospitals.
Analysis revealed that nursing workers who reported higher levels of global social support and positive social interaction/affective also reported higher frequencies of adequate practices in relation to cervical cancer (Pap smear test).In the case of the BSE, in addition to these dimensions, the emotional social support/affective were also associated with adequate practices.Thus, this study corroborates the results of other studies that point out the importance of social support in the adoption of female preventive health practices.This consistency of result attests to the positive contribution of social support to the regular practice of health care.This knowledge needs to be disseminated among nursing workers and taken into account in programs aimed to encourage female cancer screening practices.Therefore, investing in programs that value and strengthen social ties can contribute to collective and individual health promotion (9) .
Qualitative and quantitative studies aiming to clarify the specific attributes of emotional support/ information and PSI/affective and their relation with female cancer detection practices are needed.
Additionally, the identification of the more important types of social relationships that positively or negatively influence health practices (i.e.relationships with co-workers, health professionals, family members, friends, etc) can support more effective preventive strategies that take the social environment into account.

Table 2 -
Occupational characteristics of the studied

Table 3 -
Description of female cancer detection practices among studied nursing workers * categories "not informed" were excluded

Table 4 -
Prevalence and odds ratio (raw and adjusted OR) of having the Papanicolaou smear test done according to the score terciles of the social support dimensions *Adjusted for the co-variables that presented statistical association (p<0.10) in the bivariate analysis between social support and the Pap smear: age, schooling, marital status, income per capita, children, years working in nursing and weekly work hours

Table 5 -
Prevalence and odds ratio (raw and adjusted OR) of performing the BSE according to score terciles of social support dimensions *Adjusted for the co-variables that presented statistical association (p<0.10) in the bivariate analysis between social support and BSE: age, marital status, children and another job