Unmet needs of women with breast cancer during and after primary treatment: A prospective study in Denmark

Abstract

Integrated plans will be required to ensure care and rehabilitation for the growing number of cancer survivors. Information is lacking, however, about the extent to which patients’ rehabilitation needs are being met, and characteristics of patients who experience unmet needs after their diagnosis and throughout the disease trajectory. Material and methods. Consecutive women with newly diagnosed breast cancer, undergoing surgery at the Breast Surgery Clinic, Rigshospitalet, Denmark, between 2008 and 2009 (N = 261), completed questionnaires on their unmet needs, anxiety, physical functioning, social support and demographic factors at the time of surgery and four and eight months after diagnosis. Associations between demographic and clinical factors at baseline and unmet needs four and eight months after diagnosis were examined in logistic regression models. Results. The percentage of women with unmet needs remained stable between four (42%) and eight months (40%). Unmet needs were reported for patient education, counselling and alternative treatment. The factors significantly associated with having unmet needs were younger age (OR 0.92; 95% CI 0.89–0.95), higher education (OR 2.49; 95% CI 1.14–5.44), not having a partner (OR 2.25; 95% CI 1.22–4.17) and anxiety (OR 1.13; 95% CI 1.05–1.20) at four months; and age (OR 0.91; 95% CI 0.87–0.94), not having a partner (OR 2.21; 95% CI 1.10–4.46) and having had one or more unmet needs at four months (OR 6.83, 95% CI 3.55–13.16) at eight months. Conclusion. A total of 53% of women with breast cancer had unmet needs at some time between diagnosis and the end of primary treatment (eight months after diagnosis). A contextual understanding of unmet needs is necessary for planning cancer rehabilitation, as not only factors such as anxiety, physical functioning and previous unmet needs but also age, education and relationship status are associated with unmet needs.

In the growing population of cancer patients, identifying unmet needs and the characteristics of patients who develop unmet needs is essential for planning cancer care and rehabilitation [1,2]. Rehabilitation of cancer patients can be defined as ‘enabling them to reach and maintain their optimal physical, sensory, intellectual, psychological and social functional levels’ [3]. Rehabilitation is a complex process, affecting diverse aspects of patients’ lives that may differ from patient to patient, necessitating individually targeted interventions based on an understanding of each individual's current needs. Needs for rehabilitation services are here defined as those self-reported by the patients, whereas unmet needs are the needs for which no corresponding service use is reported. Unmet needs are thus defined as the discrepancy of needs and service received, similar to a definition previously employed by Houts et al. [4], however, without the specification of the degree to which the need was problematic to the patient.

In a review of studies of unmet needs among cancer patients, Harrison and colleagues reported that 1–98% of the study populations had unmet needs, which reflects the diverse approaches to identifying and defining ‘unmet needs’. Unmet needs were associated with lack of social support, low income, high age, advanced disease and lack of information on the prognosis of the disease, among other factors [5]. In two cross-sectional studies of Danish cancer patients, the prevalence of unmet needs was 10–39% (N = 1490) [6] and 17–50% (N = 3439) [7], depending on the specific need. The unmet needs most frequently reported by these samples were for physical therapy, information about non-hospital services, symptoms, late effects [6] and sexual, financial and family-oriented areas [7].

Among women with breast cancer, cross-sectional studies of small to medium-sized samples (N = 95–640) from Germany and China [8], Australia [9,10], Switzerland [11], Belgium [12] and Finland [13] have shown that medical information and communication [8,9] and fear of progression ranked highest among the needs reported [8,10,11]. Most of the studies included patients at various points in the disease trajectory [8–10], although a few limited their sample to newly diagnosed patients [11], patients in postoperative treatment [13] or patients who had completed primary treatment less than six months before the study [12]. The timing of study measurements is important, however, as needs can be expected to differ along the disease trajectory, i.e. as late effects of treatment emerge [14]. A broad range of factors was significantly associated with one or more domains of unmet needs, including age [8,10,11], education [8,11], income [10], marital status [8], anxiety, depression or distress [8,9,13], quality of life [9], physical and psychological symptom burden [8], having recurrent or metastatic disease [8,10], treatments received [8,10], nationality or nation of birth [8,11], urban or rural residence [10] and menopausal symptoms [9].

Factors associated with the use of health care services – and hence also non-use, i.e. unmet needs – can be conceptualised and structured within the ‘Behavioral Model of Health Service Use [15] (Figure 1), which explains service use in terms of three factor groups: predisposing factors, which form a dispositional basis for an individual's use of a given service; enabling factors, which mark an individual's ability to gain access to the service; and need factors, which represent the need for a given service. To understand the development of unmet needs, it may be necessary to look beyond needs to other patient characteristics.

Figure 1. The Behavioral Model of Health Service Use. Based on Andersen (1995) [15].

The aim of this prospective study was therefore to determine the incidence of unmet needs in a prospective cohort of women diagnosed and treated for primary breast cancer, surveyed at diagnosis (baseline), four months after diagnosis, when most patients are undergoing primary chemo- or radiotherapy, and eight months after diagnosis, when most patients have completed primary treatment. We also investigated baseline predisposing (age and education), enabling (having a partner and social support) and need factors (anxiety, physical functioning, high- or low-risk breast cancer and previous unmet needs) associated with unmet needs at four and eight months. Finally, we investigated having a specific unmet need for counselling by a psychiatrist or psychologist.

Material and methods

Study Design and participants

This prospective study used data collected with the primary purpose of validating the Distress Thermometer in Danish breast cancer patients [16]. Women with primary breast cancer, scheduled for surgery at the Breast Surgery Clinic of Rigshospitalet, Copenhagen, Denmark, within one week of inclusion (between 1 October 2008 and 30 October 2009) were eligible for the study (N = 426). The exclusion criteria were < 18 years of age, not residing in Denmark, no cancer or cancer in situ, relapse or secondary cancer, language or cognitive difficulties and diagnosis more than one month previously. Eligible women were invited to participate by the project nurse at the preoperation appointment the day before surgery at the Breast Surgery Clinic. Participants completed questionnaires when they were invited (baseline), four months after diagnosis (first follow-up) and eight months after diagnosis (second follow-up). Questionnaires included measures of need for and use of rehabilitation services, social support, education and anxiety.

Assessment of needs and services used

Needs for rehabilitation services were assessed in both follow-up surveys by 21 items, specifically designed for the study, covering the need to consult healthcare providers (hospital or primary care physician, hospital nurse, psychologist or psychiatrist), need for treatment other than that received in hospital (medication for psychological problems or alternative treatment), patient education (regarding the cancer, treatment, diet, relaxation or meditation, physical activity or written information about the cancer), counselling (occupational or financial, about services from patient organisations or from the local cancer rehabilitation centre), contact with other breast cancer patients or families and friends or help in dealing with practical issues (Table I). Patients were asked to indicate any need they experienced. Patients could further respond ‘don't know/not relevant’, ‘none’ and ‘other’ and specify other services needed. To assess the use of rehabilitation services at four and eight months, patients were asked to indicate whether they had used each of the services specified. Unmet needs at four and eight months were defined as needs indicated by the patients for which no corresponding rehabilitation service had been used (no unmet needs/one or more unmet needs).

Table I. Questions and items used to identify needs for rehabilitation services and services used.

1. Think about the needs for support that you have today. If you could choose, which kinds of support/treatment would you most like to use now? You may indicate several.

Consultation

With a hospital physician

With a hospital nurse

With own primary care physician

With a psychologist or psychiatrist

Treatment other than that received in hospital

Medication to treat my psychological problems

Alternative treatment (for instance by a healer)

Patient education

On my disease and treatment

On diet

On relaxation or meditation

By a physical therapist

On physical activity

Written information about the disease

Counselling

By a social worker about work or finances

Offers from patient organisations

Offers from local cancer rehabilitation centres

Contact

With other breast cancer patients in my situation

With family and friends

Other

Help in dealing with practical problems

None

Don't know or not applicable

Other

Please specify other kinds of support you would like.

2. What kinds of support or treatment did you received after your diagnosis of cancer? You may indicate several.

(The same items were repeated under this heading.)

Disease characteristics and demographic data

Information on disease characteristics, including tumour size and positive lymph node count, were extracted from the nationwide clinical database the Danish Breast Cancer Cooperative Group, which contains detailed information on nearly 95% of all breast cancer cases in Denmark since 1977 [17]. From these data, a variable indicating risk group was created: patients with tumours measuring ≤ 20 mm without lymph node involvement were classified as low risk, and those with tumours of > 20 mm or tumours of ≤ 20 mm with lymph node involvement were classified as high risk.

Age at diagnosis, relationship status (partner/no partner) and social support [having someone other than family on whom they could count for support (yes/no)] were ascertained from the baseline questionnaire. Level of formal education (basic or high school, vocational training, higher education) was derived from the four-month follow-up questionnaire.

Psychological distress and physical functioning

Anxiety was assessed at baseline using the Hospital Anxiety and Depression Scale, a 14-item scale containing separate subscales for depression and anxiety [18]. Cut-off scores can be used to identify levels of anxiety [clinical anxiety (≥ 10), no anxiety (< 8) and possible anxiety (8–10)] [18]. Physical functioning was assessed using the physical functioning scale score of the European Organization for Research and Treatment of Cancer QLQ-C30 [19]; the scale includes five items on problems experienced in physical functioning, with four response categories (‘not at all’, ‘a little’, ‘quite a bit’ and ‘very much’) and scores ranging from 0 to 100, higher scores indicating better functioning [20]. Data from the Danish Breast Cancer Cooperative Group and the questionnaires were linked by the unique personal identification number assigned to all Danish residents.

Statistical analyses

Frequencies were computed for factors that characterised the study population and for the number and type of needs reported. Logistic regression models were used to examine associations between demographic and clinical factors at baseline that were associated with having one or more unmet needs after four and eight months, respectively. Crude and mutually adjusted models were used at each follow-up to estimate odds ratios (ORs) and 95% confidence intervals (CI). At four months, age at diagnosis, breast cancer risk group, education, relationship status, social support, anxiety and physical functioning were entered into the model. At eight months, the presence of one or more unmet needs at four months was added. Separate logistic regression analyses were performed for having a specific unmet need for counselling by a psychiatrist or psychologist four and eight months after diagnosis, as an example of a specific need. The linearity of all continuous variables (age, anxiety and physical function) was assessed using linear splines [21].

Differences in anxiety and unmet needs between women who completed all three assessments (N = 311) and women who completed fewer than three assessments but for whom data on anxiety were available (N = 46) were analysed with χ²-tests. Anxiety scores were grouped into < 8, no anxiety; 8–10, possible anxiety; or > 8, clinical anxiety. All analyses were performed in SAS, version 9.

Analysed sample

Of 426 eligible patients, 367 (86%) agreed to participate, and 106 were excluded because of missing data (56 who had not completed all assessments and 50 because of missing items). The 261 patients who completed all questionnaires were included in the analyses (Figure 2).

Figure 2. Flow chart illustrating inclusion of participants in the study.

Results

The patients were aged 29–89 years at diagnosis (mean, 60); 46% had completed higher education, and the majority (68%) had a partner. At the four-month follow-up, 42% of the patients had at least one unmet need, and this number remained high (40%) at the eight-month follow-up (Table II). Persistent unmet needs were reported by 29% of patients at both the four- and the eight-month follow-up. More than half the patients (53%) had at least one unmet need during the eight months after diagnosis. Unmet needs were reported for six services; most frequently for patient-education about diet, relaxation or meditation and counselling by a psychologist or psychiatrist (Figure 3).

Table II. Demographic and clinical characteristics of 261 women with newly diagnosed breast cancer.

	No. of patients (%)
Age at diagnosis (years)	Mean, 60; SD, 10; range, 29–89
Education
Basic or high school (7–12 years)	52 (20)
Vocational training (10–12 years)	89 (34)
Higher education (≥ 13 years)	120 (46)
Partner
Yes	177 (68)
No	84 (32)
Social support outside family
Yes	228 (87)
No	33 (13)
Anxiety	Mean, 8; SD, 5; range, 0–21
Physical functioning	Mean, 92; SD, 14; range, 8–100
Risk group^a
Low risk	133 (51)
High risk	128 (49)
One or more unmet needs 4 months after diagnosis
Yes	109 (42)
No	152 (58)
One or more unmet needs 8 months after diagnosis
Yes	105 (40)
No	156 (60)
Unmet need for psychological support 4 months after diagnosis
Yes	40 (15)
No	221 (85)
Unmet need for psychological support 8 months after diagnosis
Yes	43 (16)
No	218 (84)

^aLow risk: < 20-mm tumour without lymph node involvement; high risk: ≥ 20-mm tumour or < 20-mm tumour with lymph node involvement.

Figure 3. Types of unmet needs reported by 261 female breast cancer patients 4 and 8 months after diagnosis.

Characteristics associated with unmet needs

Two factors were associated with having any unmet need both four and eight months after diagnosis in mutually adjusted analyses: younger age at diagnosis (four months: OR 0.92; 95% CI 0.89–0.95; eight months: 0.91; 0.87–0.94), and not having a partner (four months: 2.25; 1.22–4.17, eight months: 2.21; 1.10–4.46) (Table III). Higher education (2.49; 1.14–5.44) and anxiety (1.13; 1.05–1.20) were also associated with unmet needs at four months, whereas having unmet needs at four months (6.83; 3.55–13.16) was the only other factor associated with unmet needs at eight months.

Table III. Odds ratio (ORs) and 95% confidence intervals (CI) for having one or more unmet needs 4 months after diagnosis, among 261 female breast cancer patients.

	4 months after diagnosis		8 months after diagnosis
	Crude OR [95% CI]	Mutually adjusted OR [95% CI]	Crude OR [95% CI]	Mutually adjusted OR [95% CI]
Predisposing factors
Age at diagnosis (years)	0.93 [0.90–0.96]*	0.92 [0.89–0.95]*	0.91 [0.88–0.94]*	0.91 [0.87–0.94]*
Education
Basic or high school (7–12 years)	–	–	–	–
Vocational training (10–12 years)	1.14 [0.55–2.39]	1.39 [0.60–3.20]	1.27 [0.62–2.62]	1.43 [0.57–3.57]
Higher education (≥ 12 years)	2.49 [1.25–4.95]*	2.49 [1.14–5.44]*	1.68 [0.85–3.33]	0.96 [0.40–2.31]
Enabling factors
Partner
Yes	–	–	–	–
No	1.53 [0.90–2.58]	2.25 [1.22–4.17]*	1.68 [0.99–2.85]	2.21 [1.10–4.46]*
Social support
Yes	–	–	–	–
No	1.37 [0.66–2.84]	1.39 [0.90–2.13]	1.47 [0.71–3.06]	1.56 [0.96–2.53]
Need factors
Anxiety	1.12 [1.06–1.19]*	1.13 [1.05–1.20]*	1.08 [1.02–1.14]*	1.02 [0.94–1.09]
Physical functioning	1.01 [0.99–1.03]	1.00 [0.98–1.02]	1.01 [0.99–1.03]	1.00 [0.98–1.03]
Risk group^a
Low risk	–	–	–	–
High risk	1.10 [0.67–1.80]	0.94 [0.54–1.65]	1.42 [0.87–2.34]	1.36 [0.73–2.54]
One or more unmet needs at four months	–	–	9.77 [5.50–17.36]*	6.83 [3.55–13.16]*

*p < 0.05. ^aLow risk: < 20-mm tumour without lymph node involvement; high risk: ≥ 20-mm tumour or < 20-mm tumour with lymph node involvement

In the mutually adjusted analyses for an unmet need for counselling by a psychiatrist or psychologist (Table IV), younger age at diagnosis was the only factor associated with unmet needs at both four and eight months’ follow-up; risk group and anxiety were associated with unmet needs for counselling only at four months and not having a partner only at eight months.

Table IV. Odds ratios (ORs) and 95% confidence intervals (CI) for having an unmet need for counselling by a psychiatrist or psychologist 4 months after diagnosis, among 261 female breast cancer patients.

	4 months after diagnosis		8 months after diagnosis
	Crude OR [95% CI]	Mutually adjusted OR [95% CI]	Crude OR [95% CI]	Mutually adjusted OR [95% CI]
Predisposing factors
Age at diagnosis (years)	0.94 [0.91–0.97]*	0.94 [0.91–0.98]*	0.93 [0.90–0.96]*	0.93 [0.90–0.97]*
Education
Basic or high school (7–12 years)	–	–	–	–
Vocational training (10–12 years)	0.65 [0.21–2.06]	0.62 [0.18–2.13]	1.61 [0.54–4.80]	2.27 [0.65–7.97]
Higher education (≥ 12 years)	2.23 [0.86–5.77]	1.71 [0.59–4.99]	2.47 [0.89–6.87]	1.70 [0.52–5.57]
Enabling factors
Partner
Yes	–	–	–	–
No	0.89 [0.43–1.84]	1.14 [0.50–2.57]	1.47 [0.75–2.90]	2.27 [1.01–5.10]*
Social support
Yes	–	–	–	–
No	0.74 [0.24–2.22]	0.96 [0.51–1.82]	0.67 [0.22–2.01]	0.90 [0.46–1.73]
Need factors
Anxiety	1.16 [1.08–1.25]*	1.17 [1.08–1.27]*	1.11 [1.03–1.19]*	1.07 [0.98–1.17]
Physical functioning	1.03 [0.99–1.06]	1.02 [0.98–1.06]	1.01 [0.99–1.04]	0.99 [0.96–1.03]
Risk group^a
Low risk	–	–	–	–
High risk	0.51 [0.25–1.02]	0.35 [0.16–0.79]*	0.79 [0.41–1.53]	0.76 [0.34–1.67]
One or more unmet needs at four months	–	–	10.00 [4.67–21.40]*	6.95 [2.92–16.54]*

*p < 0.05. ^aLow risk: < 20-mm tumour without lymph node involvement; high risk: ≥ 20-mm tumour or < 20-mm tumour with lymph node involvement

Participants who completed fewer than three assessments

In χ²-tests comparing women who completed all assessments with those who completed fewer than three assessments, no differences were found with regard to anxiety. Patients for whom the data were complete were significantly more likely to have one or more unmet needs at both follow-up assessments (p < 0.001).

Discussion

On average, 41% of our sample of 261 women with a diagnosis of primary breast cancer expressed unmet needs four and eight months after diagnosis. These women were younger, had a higher educational level, had no partner, expressed greater anxiety and were more likely to have unmet needs at both follow-up assessments than those not experiencing unmet needs. Despite some methodological differences that make direct comparisons difficult, our results are in line with those of two cross-sectional studies from Switzerland [11] and Belgium [12], in which the prevalence of unmet needs was 36–40% for women with newly diagnosed breast cancer [11] and 28–59% for breast cancer patients who had completed primary treatment [12]. Whereas Pauwels and colleagues assessed needs using single items reflecting broad themes (such as psychological or physical functioning or sexuality) [12], Griesser and colleagues used the Supportive Care Needs Survey [11], and our assessment was based on single items covering specific services. The different assessment methods also imply conceptual differences: needs assessments were based either on services that have not been received or on the problems for which patients did not receive adequate services. We assessed both self-reported needs and services used and combined them into a measure of unmet needs, based on the discrepancy between needs and services. Needs were self-reported and thus ‘perceived’ by patients; however, they were judged as met or unmet by the researchers (and not by the patients) and can therefore be characterised as ‘evaluated’. In general, the way in which needs and unmet needs are defined may be ambiguous; i.e. as some studies assess whether needs are present and whether they are met by the same questions simultaneously [10], limiting comparability.

Characteristics associated with unmet needs

While education, age, relationship status and anxiety were associated with unmet needs at four months, only age, relationship status and unmet needs at four months were significantly associated with unmet needs at eight months. When the factors that we found to be significantly associated with unmet needs are viewed within the Behavioral Model of Health Service Use, all factors groups (predisposing, enabling and need) play a role for developing unmet needs at four and eight months after diagnosis. This indicates that it is necessary to look beyond needs (in terms of symptoms), when understanding unmet needs among patients both during and after primary treatment. This also applies to the specific unmet need for counselling by a psychiatrist or psychologist after primary treatment.

We found that younger patients were more likely to have unmet needs, in line with the findings of previous studies [8,10,11]. Part of the effect of younger age on unmet needs could be carried by distress, as younger age was previously related to distress in this sample [22], and distress was significantly associated with self-reported inadequate psychosocial treatment in breast cancer patients [13]. We did adjust for anxiety using the HADS anxiety measure and the remaining effect of age may thus reflect residual confounding from distress. Further, we found that education was significantly associated with unmet needs at the earlier follow-up. Griesser et al. also found that higher education was significantly associated with unmet needs [11], as did Lam et al. [8]; however, Luutonen et al. found the opposite association [13]. Higher education might influence unmet needs by way of higher expectations for rehabilitation services [6]. Patients who had a partner were less likely to have unmet needs in this study, suggesting that social support from partners may lower need or helps patients to identify or access services [23].

As needs for different rehabilitation services have been found to be associated with varying patient characteristics [24], the factors associated with different unmet needs might also vary. In analyses of patients who expressed the need for a consultation with a psychiatrist or psychologist, younger age, anxiety and having low-risk breast cancer were associated with this need after four months, whereas age, not having a partner and any needs at four months were associated with this need after eight months. The association between age and the need for counselling might be due to the higher prevalence of distress among younger patients [22]. As we did adjust for anxiety the remaining effect of age may reflect residual confounding from distress. The association with not having a partner could indicate an effect of lack of support [23]. The association between having low-risk breast cancer and having an unmet need for counselling might appear to be counterintuitive; however, patients with high-risk breast cancer might more readily be offered counselling than low-risk patients, who would consequently not have their counselling needs met. Younger age and worse mental health were associated with use of subsidised psychotherapy in a sample of Danish cancer patients [25], showing that patients with the greatest need do indeed attend counselling. These results suggest that the factors related to having any unmet need and those related to having a specific unmet need may differ, but that the associations are nevertheless strikingly similar. Therefore, the development of unmet needs might depend on general rather than specific service-related factors.

Strengths and limitations

The longitudinal design of this study is an advantage in comparison with the previous cross-sectional studies [8–12], as it allowed us to follow patients prospectively for eight months from diagnosis, and thus investigate factors that preceded the development of unmet needs temporally. Further, to the best of our knowledge, this is the first prospective study enrolling participants at the same time in the disease trajectory, as all patients were included fewer than four weeks after diagnosis, allowing for the investigation of factors related to unmet needs at specific times, during and after active treatment. We had a high response rate at baseline (86%), implying that the results can be generalised; however, not all participants completed all three assessments, and analyses of participants who did not complete all assessments and were excluded from the analyses showed that these patients were less likely to have unmet needs. The estimates may therefore be higher than the true prevalence, thus limiting the generalisability of the findings. The associations of predisposing, enabling and need factors with unmet needs are not, however, expected to be affected by this bias. Due to the sample size and limited power, analyses of interactions among patient characteristics were not performed, which constitutes a limitation of this study. The assessment of unmet needs in this study might have introduced some misclassification, as need for and use of a given service might have overlapped. This misclassification would, however, tend to underestimate the prevalence of unmet needs. Further, our method for assessing needs is not validated, and comparison with existing measures is therefore limited, e.g. by not assessing compiled domains of needs, but individual services instead. Nevertheless, the straightforward character of the items makes them easy to interpret in terms of the prevalence of needs. For instance, the Supportive Care Needs Survey [26] could be used in future studies of Danish cancer patients to ensure comparability across studies. The medium-sized sample probably limited the statistical power of the study, but the results contribute to the understanding of unmet needs in newly diagnosed breast cancer patients.

More than half (53%) of patients with primary, incident breast cancer expressed unmet needs within the first eight months of diagnosis, and such unmet needs should be understood in a contextual perspective, including not only symptoms that indicate need but also predisposing and enabling factors.

Conclusion

If our results are confirmed, clinicians should be aware that young women with breast cancer who do not have partners may be more prone to experience unmet needs for rehabilitation services, from the time of diagnosis and throughout primary treatment.

Acknowledgements

This study was supported by the Nordic Cancer Union and the Committee for Psychosocial Cancer Research at the Danish Cancer Society.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.