Support Care Cancer DOI 10.1007/s00520-015-2836-0
ORIGINAL ARTICLE
Identification of health care needs of long-term breast cancer survivors among Israeli women Lori Mandelzweig 1 & Angela Chetrit 1 & Tova Amitai 1 & Bernice Oberman 1 & Nava Siegelmann Danieli 2 & Barbara Silverman 3 & Siegal Sadetzki 1,4
Received: 3 March 2015 / Accepted: 22 June 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose Improvement in treatment has extended survival of breast cancer patients. Our study aimed to characterize health service use among long-term breast cancer survivors in Israel in order to identify and address specific needs of this subpopulation. Methods The study population included 250 women diagnosed with breast cancer, 8–12 years prior to study initiation (cases), and 250 individually matched cancer-free controls. Participants were recruited from the second largest Israeli HMO, and data were collected through personal interviews. ORs and 95 % CIs were estimated using conditional logistic regression and generalized estimating equations. Results Greater use of health services was observed among cases, compared to an age-matched comparison group, expressed by more visits to family physicians and specialists, longer duration of visits, more requests for referrals, more frequent contact with emergency services, and hospitalizations. The study groups were similar regarding socioeconomic variables, current smoking and physical activity, BMI, and prevalence of reported morbidity. Although 80 % of cases defined the family physician as their main treating physician, half still considered their oncologist responsible for cancer follow-up. Only 14.4 and 10.4 % of cases and controls, * Siegal Sadetzki [email protected] 1
Cancer & Radiation Epidemiology Unit, Gertner Institute for Epidemiology & Health Policy Research, Chaim Sheba Medical Center, Tel Hashomer, Ramat Gan 5262000, Israel
2
Maccabi Health Services, Tel Aviv, Israel
3
Israel National Cancer Registry, Jerusalem, Israel
4
Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
respectively, reported receiving psychological support during the year preceding the interview. Conclusions Further studies should assess the contribution of apprehension concerning health-related issues that still accompany breast cancer survivors, to the excess use of health services. Concern among family practitioners may contribute as well. In addition, our results emphasize the need to improve coordination between the disciplines of oncology and community medicine for the medical care of this group. Keywords Breast cancer . Cancer survivors . Health service use . Care needs . Satisfaction with health services
Introduction Significant improvements in diagnosis and treatment of breast cancer (BC), resulting from early detection and efficacy of new treatments, have contributed to better prognosis of the disease [1]. The current number of BC survivors in the USA has been estimated at more than 3.1 million [2]. According to statistics published by the American Cancer Society in 2014, 61 % of BC cases are diagnosed at a localized stage, with a 5-year relative survival rate of 99 %. For all stages combined, relative survival rates of 83 and 78 % at 10 and 15 years after diagnosis, respectively, were reported [3]. In Israel as well, improvement in relative 5-year survival rates among women diagnosed with BC in 2001–2006 has been observed, with rates of 87.2 and 78.9 % for Israeli and Arab women, respectively, compared with rates of 78.1 and 67 %, respectively, among those diagnosed between 1991 and 1995 [4]. Consequently, long periods of survival following completion of treatment have become the norm for many BC survivors. Over the years, interest in cancer survivorship has increased, and BC has become “a curable disease for some and a chronic disease for others” [5].
Support Care Cancer
These changes have been accompanied by a need for surveillance for recurrence, screening, and preventive care for other diseases. Other medical issues requiring attention include late and long-term effects of treatment such as infertility, menopausal symptoms, lymphedema, fatigue, osteoporosis, sleep disturbances, neuropathy, cognitive issues, weight gain, pain, cardiac dysfunction, sexual dysfunction, and second cancers [6–9]. Another important aspect of long-term cancer survivorship that must be addressed is the aging of the survivor population, accompanied by increases in comorbidity. In the case of BC, SEER data (USA) show that 40 % of affected women are aged 65 or over when diagnosed (median 61) [10]. Consequently, many women surviving 10 years are likely to be in their 70s. It has been suggested that among older survivors, non-cancer comorbidities may affect functional status, quality of life, and survival more than their previous cancer [11]. However, there is a concern that cancer survivors may not be receiving the appropriate care and attention needed for non-cancer conditions, due to the focus on issues related to cancer or because of a lack of clarity concerning the roles of oncologists and primary care providers [11]. Beyond medical issues, BC survivors have informational and emotional needs, which, when fulfilled, may contribute to their well-being [12]. A recent survey conducted at the Mayo Clinic among BC survivors showed that 73 % strongly agreed that their medical needs were met, while only 49 % strongly agreed that their psychological and spiritual needs were being met [13]. The overall goals of this study were to characterize health service use among long-term BC survivors in Israel and to identify areas within the health care system which require improvement in order to address the needs of this population. The specific aim of the study was to compare health status, health service use, and satisfaction with health services between women diagnosed with BC 8–12 years in the past and a comparable group of women with no history of cancer.
Materials and methods The study participants lived in central Israel and were members of Maccabi Healthcare Services, an HMO which insures 25 % of the Israeli population. The cases included women diagnosed with BC between 1999 and 2003, who had no history of cancer prior to BC diagnosis and had no evidence of the disease after 8–12 years. The comparison group included women with no history of any cancer, individually matched to cases by age (±2 years) and area of residence. Cases and controls who were found to have another cancer, recurrence of BC, cognitive impairment, language issues, or died were excluded from the study.
Overall, in the defined geographic area of the study, 596 cases complied with the eligibility criteria. Figure 1 describes the process of recruitment of the study population. Up to 10 potential controls were randomly selected for each case. Two hundred and fifty cases were randomly selected and approached. Each woman who did not agree to participate, could not be contacted after several attempts, or was found not eligible for the study according to the abovementioned exclusion criteria, was replaced by the consecutive women among the remaining eligible subjects. By the end of the recruitment process, all potential cases who lived in the geographic study area were contacted. For each case that was successfully enrolled, a control was approached resulting in 696 controls who were contacted. One hundred and eighteen cases and 114 controls were excluded resulting in 478 cases and 582 comparison group members of which 250 women were interviewed in each group (response rates of 52.3 and 43 %, respectively). The difference in age between those interviewed and those not interviewed was not statistically significant among the total group or for cases and controls separately. Data were collected in face-to-face interviews by experienced interviewers trained in workshops. A structured questionnaire was used for data collection and included sociodemographic questions, lifestyle habits, morbidity, and health service use (frequency, reasons, and duration of visits to family physicians and specialists; receipt of psychological support; contact with medical emergency services; and hospitalizations). The questionnaire also included questions on satisfaction with health care services. Prior to the conduction of interviews, participants signed informed consent forms. The study protocol was approved by the Institutional Review Boards of the Sheba Medical Center and Maccabi Healthcare Services. Statistical methods Several dependent variables were analyzed in this study, including frequency of visits to family physicians and specialists, number of visits by reasons for visits, average duration of visits, receipt of psychological support from a psychologist or social worker, contact with medical emergency services, and hospitalizations. The covariates included status of the participant (case/comparison group, main variable), sociodemographic variables, level of religious observance, number of children, smoking, physical activity, and body mass index. The level of leisure physical activity was measured in metabolic equivalent of task (MET) scores and was categorized into four groups (no physical activity and tertiles of MET hours). In addition, self-reported information on morbidity was collected for selected groups of diseases. Differences in independent variables between BC survivors and controls were examined using the chi-square test or Fisher’s exact test, when needed, for categorical variables; t tests or the Kruskal-Wallis non-parametric test was used for
Support Care Cancer Fig. 1 Description of recruitment of the study population
Cases
Controls
Target populaon:
N=596
Up to 10 random controls/case
Approached:
N=596
N=696
Excluded N=118
continuous variables. Conditional logistic regression analyses were applied for dichotomous dependent outcomes. Differences between cases and the comparison group for the extent of health service use (number of visits to family physicians and specialists) were assessed using generalized estimating equations (GEE). Satisfaction was assessed by ratings of the quality of doctor-patient communication and the patient’s perception of the extent to which the treating physician was updated concerning treatment provided to them by other health professionals, using the GPAS-2—General Practice Assessment Survey: You and Your Doctor [14].
Results The mean age of the study participants at the time of interview was 65.0±10.3 years. Socioeconomic characteristics and lifestyle factors are presented in Table 1. No significant differences in the distributions of origin by continent, marital status, level of religiosity, level of education, and occupation were found between the groups. While distributions of level of education in the study groups were similar, a greater percentage of cases compared to the comparison group finished their education at the level of primary school only. In comparison with the average monthly income in Israel, 18 and 13 % of cases and the comparison group, respectively, reported incomes much lower than average (p=0.10). While fewer cases were nulliparous, more members of the comparison group had three or more children. Current smoking, the distribution of BMI by categories (underweight, normal weight, overweight, and obese), and the distribution of physical activity levels were similar in the study groups. Mean BMIs were 26.0±4.4 and 25.9±4.4 and median MET scores at the time of interview were 11.9 and 15.3 for cases and comparison group members, respectively (not shown).
Eligible N=478
Excluded N=114
Eligible N=582
Interviewed N=250
Interviewed N=250
(response rate 52.3%)
(response rate 43%)
Analysis of self-reported morbidity showed that the highest frequency of diseases among the study population were those related to the endocrine and circulatory systems. The only statistically significant difference observed between the study groups in terms of morbidity was in the urinary system, where 8.4 % of cases and 3.2 % of the controls reported diseases. Urinary tract infection was the most prevalent disease of this system, followed by calculus of the kidney (Table 2). Table 3 presents the annual consultation rates for general practitioners and a variety of specialists, as well as annual contact with emergency services, visits to psychologists, and hospitalizations by study group. More than 90 % of the study participants visited family physicians at least once during the past year. Over the last year, specialists were visited by 84 % of the study population, three times more by cases compared to the comparison group (95 % CI 1.78–5.49). Cases visited surgeons and gynecologists more than members of the comparison group (OR 2.82; 95 % CI 1.81–4.37 and OR 1.68; 95 % CI 1.10–2.56, respectively). Only 14 % of cases and 10 % of the comparison group reported receiving psychological support over the past year (OR 1.55; 95 % CI 0.88–2.72). Non-statistically significant elevated odds ratios were seen for contact with an ambulance or emergency health service and for being hospitalized at least once during the past 5 years. A total of 1758 visits to the family physician were reported over the last year, with the median number of visits being greater among cases compared to the comparison group (4 versus 3, p=0.02). The probability of a BC survivor having a greater number of visits to a family physician in the year preceding the interview was, on average, 50 % higher than that of age-matched members of the comparison group (95 % CI 1.07–2.06) (Table 4). The most frequent reason for visiting the family physician (occurring in more than 40 %) among both groups
Support Care Cancer Table 1 Description of the study population by socioeconomic characteristics and lifestyle factors
Characteristics
Origina
Family status
Level of religiosity
Education
Income
Occupation
Number of children
Current smoker Current BMI (kg/m2)
Current PA level
Category
Cases
Controls
p value
N
%
N
%
Asia Africa
40 28
16.0 11.2
46 24
18.4 9.6
Europe/America Israel
138 13
55.2 5.2
136 16
54.4 6.4
Mixed Married/living with partner Single Divorced/separated/widow
31 158 14 78
12.4 63.2 5.6 31.2
28 148 17 85
11.2 59.2 6.8 34.0
Secular Traditional
154 64
61.6 25.6
148 77
59.4 30.9
Observant/ultraorthodox Primary school High school
32 18 74
12.8 7.2 29.6
24 7 89
9.6 2.8 35.6
Post-high school non-academic Bachelor’s degree Master’s degree or higher Higher than average income Similar to average income Lower than average income
53 61 43 59 49 53
21.2 24.4 17.2 23.6 19.6 21.2
50 56 45 74 48 55
20.0 22.4 18.0 29.6 19.2 22.0
Much lower than average income Unknown
44 45
17.6 18.0
32 41
12.8 16.4
White collar worker Blue collar worker Housewife Unknown 0 1–2 3+ Yes No
124 101 17 8 17 120 113 39 211
49.6 40.4 6.8 3.2 6.8 48.0 45.2 15.6 84.4
116 117 13 4 29 96 125 33 217
46.4 46.8 5.2 1.6 11.6 38.4 50.0 13.2 86.8
0.4
Underweight (
ORIGINAL ARTICLE
Identification of health care needs of long-term breast cancer survivors among Israeli women Lori Mandelzweig 1 & Angela Chetrit 1 & Tova Amitai 1 & Bernice Oberman 1 & Nava Siegelmann Danieli 2 & Barbara Silverman 3 & Siegal Sadetzki 1,4
Received: 3 March 2015 / Accepted: 22 June 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose Improvement in treatment has extended survival of breast cancer patients. Our study aimed to characterize health service use among long-term breast cancer survivors in Israel in order to identify and address specific needs of this subpopulation. Methods The study population included 250 women diagnosed with breast cancer, 8–12 years prior to study initiation (cases), and 250 individually matched cancer-free controls. Participants were recruited from the second largest Israeli HMO, and data were collected through personal interviews. ORs and 95 % CIs were estimated using conditional logistic regression and generalized estimating equations. Results Greater use of health services was observed among cases, compared to an age-matched comparison group, expressed by more visits to family physicians and specialists, longer duration of visits, more requests for referrals, more frequent contact with emergency services, and hospitalizations. The study groups were similar regarding socioeconomic variables, current smoking and physical activity, BMI, and prevalence of reported morbidity. Although 80 % of cases defined the family physician as their main treating physician, half still considered their oncologist responsible for cancer follow-up. Only 14.4 and 10.4 % of cases and controls, * Siegal Sadetzki [email protected] 1
Cancer & Radiation Epidemiology Unit, Gertner Institute for Epidemiology & Health Policy Research, Chaim Sheba Medical Center, Tel Hashomer, Ramat Gan 5262000, Israel
2
Maccabi Health Services, Tel Aviv, Israel
3
Israel National Cancer Registry, Jerusalem, Israel
4
Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
respectively, reported receiving psychological support during the year preceding the interview. Conclusions Further studies should assess the contribution of apprehension concerning health-related issues that still accompany breast cancer survivors, to the excess use of health services. Concern among family practitioners may contribute as well. In addition, our results emphasize the need to improve coordination between the disciplines of oncology and community medicine for the medical care of this group. Keywords Breast cancer . Cancer survivors . Health service use . Care needs . Satisfaction with health services
Introduction Significant improvements in diagnosis and treatment of breast cancer (BC), resulting from early detection and efficacy of new treatments, have contributed to better prognosis of the disease [1]. The current number of BC survivors in the USA has been estimated at more than 3.1 million [2]. According to statistics published by the American Cancer Society in 2014, 61 % of BC cases are diagnosed at a localized stage, with a 5-year relative survival rate of 99 %. For all stages combined, relative survival rates of 83 and 78 % at 10 and 15 years after diagnosis, respectively, were reported [3]. In Israel as well, improvement in relative 5-year survival rates among women diagnosed with BC in 2001–2006 has been observed, with rates of 87.2 and 78.9 % for Israeli and Arab women, respectively, compared with rates of 78.1 and 67 %, respectively, among those diagnosed between 1991 and 1995 [4]. Consequently, long periods of survival following completion of treatment have become the norm for many BC survivors. Over the years, interest in cancer survivorship has increased, and BC has become “a curable disease for some and a chronic disease for others” [5].
Support Care Cancer
These changes have been accompanied by a need for surveillance for recurrence, screening, and preventive care for other diseases. Other medical issues requiring attention include late and long-term effects of treatment such as infertility, menopausal symptoms, lymphedema, fatigue, osteoporosis, sleep disturbances, neuropathy, cognitive issues, weight gain, pain, cardiac dysfunction, sexual dysfunction, and second cancers [6–9]. Another important aspect of long-term cancer survivorship that must be addressed is the aging of the survivor population, accompanied by increases in comorbidity. In the case of BC, SEER data (USA) show that 40 % of affected women are aged 65 or over when diagnosed (median 61) [10]. Consequently, many women surviving 10 years are likely to be in their 70s. It has been suggested that among older survivors, non-cancer comorbidities may affect functional status, quality of life, and survival more than their previous cancer [11]. However, there is a concern that cancer survivors may not be receiving the appropriate care and attention needed for non-cancer conditions, due to the focus on issues related to cancer or because of a lack of clarity concerning the roles of oncologists and primary care providers [11]. Beyond medical issues, BC survivors have informational and emotional needs, which, when fulfilled, may contribute to their well-being [12]. A recent survey conducted at the Mayo Clinic among BC survivors showed that 73 % strongly agreed that their medical needs were met, while only 49 % strongly agreed that their psychological and spiritual needs were being met [13]. The overall goals of this study were to characterize health service use among long-term BC survivors in Israel and to identify areas within the health care system which require improvement in order to address the needs of this population. The specific aim of the study was to compare health status, health service use, and satisfaction with health services between women diagnosed with BC 8–12 years in the past and a comparable group of women with no history of cancer.
Materials and methods The study participants lived in central Israel and were members of Maccabi Healthcare Services, an HMO which insures 25 % of the Israeli population. The cases included women diagnosed with BC between 1999 and 2003, who had no history of cancer prior to BC diagnosis and had no evidence of the disease after 8–12 years. The comparison group included women with no history of any cancer, individually matched to cases by age (±2 years) and area of residence. Cases and controls who were found to have another cancer, recurrence of BC, cognitive impairment, language issues, or died were excluded from the study.
Overall, in the defined geographic area of the study, 596 cases complied with the eligibility criteria. Figure 1 describes the process of recruitment of the study population. Up to 10 potential controls were randomly selected for each case. Two hundred and fifty cases were randomly selected and approached. Each woman who did not agree to participate, could not be contacted after several attempts, or was found not eligible for the study according to the abovementioned exclusion criteria, was replaced by the consecutive women among the remaining eligible subjects. By the end of the recruitment process, all potential cases who lived in the geographic study area were contacted. For each case that was successfully enrolled, a control was approached resulting in 696 controls who were contacted. One hundred and eighteen cases and 114 controls were excluded resulting in 478 cases and 582 comparison group members of which 250 women were interviewed in each group (response rates of 52.3 and 43 %, respectively). The difference in age between those interviewed and those not interviewed was not statistically significant among the total group or for cases and controls separately. Data were collected in face-to-face interviews by experienced interviewers trained in workshops. A structured questionnaire was used for data collection and included sociodemographic questions, lifestyle habits, morbidity, and health service use (frequency, reasons, and duration of visits to family physicians and specialists; receipt of psychological support; contact with medical emergency services; and hospitalizations). The questionnaire also included questions on satisfaction with health care services. Prior to the conduction of interviews, participants signed informed consent forms. The study protocol was approved by the Institutional Review Boards of the Sheba Medical Center and Maccabi Healthcare Services. Statistical methods Several dependent variables were analyzed in this study, including frequency of visits to family physicians and specialists, number of visits by reasons for visits, average duration of visits, receipt of psychological support from a psychologist or social worker, contact with medical emergency services, and hospitalizations. The covariates included status of the participant (case/comparison group, main variable), sociodemographic variables, level of religious observance, number of children, smoking, physical activity, and body mass index. The level of leisure physical activity was measured in metabolic equivalent of task (MET) scores and was categorized into four groups (no physical activity and tertiles of MET hours). In addition, self-reported information on morbidity was collected for selected groups of diseases. Differences in independent variables between BC survivors and controls were examined using the chi-square test or Fisher’s exact test, when needed, for categorical variables; t tests or the Kruskal-Wallis non-parametric test was used for
Support Care Cancer Fig. 1 Description of recruitment of the study population
Cases
Controls
Target populaon:
N=596
Up to 10 random controls/case
Approached:
N=596
N=696
Excluded N=118
continuous variables. Conditional logistic regression analyses were applied for dichotomous dependent outcomes. Differences between cases and the comparison group for the extent of health service use (number of visits to family physicians and specialists) were assessed using generalized estimating equations (GEE). Satisfaction was assessed by ratings of the quality of doctor-patient communication and the patient’s perception of the extent to which the treating physician was updated concerning treatment provided to them by other health professionals, using the GPAS-2—General Practice Assessment Survey: You and Your Doctor [14].
Results The mean age of the study participants at the time of interview was 65.0±10.3 years. Socioeconomic characteristics and lifestyle factors are presented in Table 1. No significant differences in the distributions of origin by continent, marital status, level of religiosity, level of education, and occupation were found between the groups. While distributions of level of education in the study groups were similar, a greater percentage of cases compared to the comparison group finished their education at the level of primary school only. In comparison with the average monthly income in Israel, 18 and 13 % of cases and the comparison group, respectively, reported incomes much lower than average (p=0.10). While fewer cases were nulliparous, more members of the comparison group had three or more children. Current smoking, the distribution of BMI by categories (underweight, normal weight, overweight, and obese), and the distribution of physical activity levels were similar in the study groups. Mean BMIs were 26.0±4.4 and 25.9±4.4 and median MET scores at the time of interview were 11.9 and 15.3 for cases and comparison group members, respectively (not shown).
Eligible N=478
Excluded N=114
Eligible N=582
Interviewed N=250
Interviewed N=250
(response rate 52.3%)
(response rate 43%)
Analysis of self-reported morbidity showed that the highest frequency of diseases among the study population were those related to the endocrine and circulatory systems. The only statistically significant difference observed between the study groups in terms of morbidity was in the urinary system, where 8.4 % of cases and 3.2 % of the controls reported diseases. Urinary tract infection was the most prevalent disease of this system, followed by calculus of the kidney (Table 2). Table 3 presents the annual consultation rates for general practitioners and a variety of specialists, as well as annual contact with emergency services, visits to psychologists, and hospitalizations by study group. More than 90 % of the study participants visited family physicians at least once during the past year. Over the last year, specialists were visited by 84 % of the study population, three times more by cases compared to the comparison group (95 % CI 1.78–5.49). Cases visited surgeons and gynecologists more than members of the comparison group (OR 2.82; 95 % CI 1.81–4.37 and OR 1.68; 95 % CI 1.10–2.56, respectively). Only 14 % of cases and 10 % of the comparison group reported receiving psychological support over the past year (OR 1.55; 95 % CI 0.88–2.72). Non-statistically significant elevated odds ratios were seen for contact with an ambulance or emergency health service and for being hospitalized at least once during the past 5 years. A total of 1758 visits to the family physician were reported over the last year, with the median number of visits being greater among cases compared to the comparison group (4 versus 3, p=0.02). The probability of a BC survivor having a greater number of visits to a family physician in the year preceding the interview was, on average, 50 % higher than that of age-matched members of the comparison group (95 % CI 1.07–2.06) (Table 4). The most frequent reason for visiting the family physician (occurring in more than 40 %) among both groups
Support Care Cancer Table 1 Description of the study population by socioeconomic characteristics and lifestyle factors
Characteristics
Origina
Family status
Level of religiosity
Education
Income
Occupation
Number of children
Current smoker Current BMI (kg/m2)
Current PA level
Category
Cases
Controls
p value
N
%
N
%
Asia Africa
40 28
16.0 11.2
46 24
18.4 9.6
Europe/America Israel
138 13
55.2 5.2
136 16
54.4 6.4
Mixed Married/living with partner Single Divorced/separated/widow
31 158 14 78
12.4 63.2 5.6 31.2
28 148 17 85
11.2 59.2 6.8 34.0
Secular Traditional
154 64
61.6 25.6
148 77
59.4 30.9
Observant/ultraorthodox Primary school High school
32 18 74
12.8 7.2 29.6
24 7 89
9.6 2.8 35.6
Post-high school non-academic Bachelor’s degree Master’s degree or higher Higher than average income Similar to average income Lower than average income
53 61 43 59 49 53
21.2 24.4 17.2 23.6 19.6 21.2
50 56 45 74 48 55
20.0 22.4 18.0 29.6 19.2 22.0
Much lower than average income Unknown
44 45
17.6 18.0
32 41
12.8 16.4
White collar worker Blue collar worker Housewife Unknown 0 1–2 3+ Yes No
124 101 17 8 17 120 113 39 211
49.6 40.4 6.8 3.2 6.8 48.0 45.2 15.6 84.4
116 117 13 4 29 96 125 33 217
46.4 46.8 5.2 1.6 11.6 38.4 50.0 13.2 86.8
0.4
Underweight (
