Preferences for Pharmacist Counselling in Patients with Breast Cancer: A Discrete Choice Experiment

Abstract

With the shift of a large proportion of cancer chemotherapy recipients to ambulatory care, the role of hospital pharmacists has changed, and their provision of information is essential care for cancer patients. There is little research on pharmacist–patient relations, particularly about pharmacist counselling, in Japan. To meet patients’ needs, pharmacist counselling should be optimized. Here, breast cancer patients’ preferences for pharmacist counselling were assessed using a discrete choice experiment. Bayesian nonlinear optimal methodology was employed to obtain six attributes (attitude of pharmacist, quality of information, explanation of side effects, frequency of pharmacist counselling before starting chemotherapy, cost of pharmacist counselling, and follow-up with the pharmacist after starting chemotherapy) of two to three levels each. The attributes and levels were used to create 12 hypothetical scenarios that were divided into two questionnaires of six choice sets each. Two hundred eighty participants were randomly assigned to complete one of these questionnaires (blocks). Attributes were analyzed by conditional logit model to determine significant predictors of patient preferences. The responses of 278 patients to 1667 scenarios were analyzed. Attitude of pharmacist, quality of information, cost of pharmacist counselling, and follow-up with the pharmacist after starting chemotherapy were significant predictors of patient preferences, with quality of information receiving the highest priority. Thus patients receiving pharmacist counselling before starting chemotherapy prefer to interact with a pharmacist with a friendly, interested attitude who provides individualized information. Further research is needed to elucidate the information that Japanese patients consider most important and to enhance pharmacist–patient communication.

Breast cancer is the leading type of cancer in Japanese women, and the estimated number of newly diagnosed breast cancer patients was approximately 56000 in 2007.¹⁾ The majority of breast cancer patients receive chemotherapy in outpatient settings and have problematic symptoms including insomnia, oral problems, appetite loss, nausea, fatigue, and pain.^2,3) These possible side effects vary by chemotherapeutic regimen. The expected benefits and possible side effects cause decisional conflicts about treatment choice for patients. Before starting chemotherapy, some patients exhibit high decisional conflict regarding treatment choice, including the option of no treatment.⁴⁾ It has been shown that pharmacist counselling is able to decrease this decisional conflict,⁴⁾ and more than 75% of patients express interest in pharmacist counselling about their medication and follow-up with the pharmacist during chemotherapy.⁵⁾ Gourdji et al. found that patients with cancer prefer to interact with pharmacists regarding their medication as part of outpatient services.⁶⁾ This provision of information is an important pharmacist responsibility and is part of essential care for cancer patients.^7–9) Provision of sufficient information benefits patients through shared decision-making, better communication, and increased quality of life.^8–10)

The practice of hospital pharmacists differs from country to country, but in general, oncology pharmacists are now expected to be competent in both drug-oriented services and patient-centred care.^11,12) As the roles of pharmacists change, there is growing interest in the conceptualization and measurement of pharmacist–patient communication. The pharmacist–patient relationship has been conceptualized as a transmission model and a transactional model.¹³⁾ The former is a one-way process model of biomedical communication in which pharmacists provide information mainly about medical issues. The latter describes a two-way process of patient-centred communication in which pharmacist counselling considers the patient’s perspective, with special attention paid to psychological or social aspects. In addition to being able to provide patient-centred care, hospital pharmacists are also expected to provide pharmaceutical care in inpatient settings and in outpatient chemotherapy settings.¹⁴⁾ In 2005, the Japanese Society of Hospital Pharmacists established certifications for Oncology Pharmacy Specialists.¹⁵⁾

To optimize pharmacist counselling, it is essential to understand the patient’s perspective. Therefore, this study aimed to explore patient preferences for pharmacist counselling by using a discrete choice experiment (DCE). A DCE is a stated preference method in which respondents are asked to make choices about hypothetical situations presented by the researcher. Use of DCEs was originally developed in economics^16,17) and they have been widely implemented since the early 1990 s to elicit patients’ preferences about health care services and interventions.¹⁸⁾ According to a review by Naik-Panvelkar et al., DCE studies about pharmacy services have been conducted only in North America and Europe.¹⁹⁾ No DCE studies, especially in hospital pharmacy settings, have been conducted in Japan or Asia. In addition, although the paradigm shifts described above have been occurring worldwide, little research has been conducted about the pharmacist–patient relationship in Japan. The primary goal of this study was to determine the components of pharmacist counselling that patients consider important.

MATERIALS AND METHODS

Ethical Considerations

The protocol, including both the pilot and main studies, was approved by the institutional review board of Tokyo Medical University and Nagumo Clinic. This study was conducted in accordance with Japanese ethical guidelines for epidemiological research and the Declaration of Helsinki. All patients gave written informed consent. This study was registered in the University Hospital Medical Information Network Clinical Trials Registry (trial number: 000008599).

Study Design

A stated preference DCE using forced choice format was designed to deduce patient preferences for pharmacist counselling. The necessity of additional options in the questionnaire, such as opt-out or status quo, is controversial and depends on its design. These options may be inadequate for research in health care.²⁰⁾ Considering the actual current practice of pharmacist counselling in Japan, patients ordinarily do not avoid pharmacist counselling. Thus, we pragmatically decided not to use opt-out or status quo as additional options. The choice set in a DCE can be created through linear or nonlinear methods. The classical linear method uses full and fractional factorial design and orthogonal arrays to create the choice set, but this method may result in DCE parameters that do not fit the linear models.²¹⁾ A nonlinear method for designing the choice set is utilization of Bayesian optimal designs. Kessels and colleagues reported that the Bayesian optimal design methodology is more efficient than the linear methodology.²¹⁾ Bayesian optimal designs can be generated algorithmically by using JMP statistical software (version 9.02, SAS Institute Inc., Cary, NC, U.S.A.). For this study, we utilized JMP to create the DCE using Bayesian nonlinear optimal design.

Attributes and Levels

No DCE studies on pharmacist counselling in ambulatory care settings have been conducted, but several conceptually similar DCE studies investigating other types of counselling were found.^22–24) In addition to the attributes of counselling examined in these previous studies, we developed a list of other possible attributes: frequency of pharmacist counselling, place of counselling, time of counselling, presence of a primary care pharmacist, treatment description, cost of counselling, and information about effects of chemotherapy on sexual activity. We conducted individual interviews with two breast cancer patients (who met the study criteria) regarding patient preference for pharmacist counselling before starting chemotherapy. The president of a professional organization of hospital pharmacists, a board certified oncology pharmacy specialist, a researcher, a methodologist, and a breast cancer survivor reviewed the interview content and results of literature review, and reached consensus on the selection of attributes and levels. The six attributes are (1) ATTITUDE of pharmacist, (2) Quality of INFORMATION, (3) Explanation of SIDE EFFECTS, (4) FREQUENCY of pharmacist counselling before starting chemotherapy, (5) COST of pharmacist counselling, and (6) FOLLOW-UP with the pharmacist after starting chemotherapy. The levels of each attribute are shown in Table 1. The levels for ATTITUDE, INFORMATION, and FOLLOW-UP were determined using the report of Empel et al. as a ref. 24. The levels for SIDE EFFECTS focused on the main content of information and were determined according to prior studies.^25,26) The levels for cost of pharmacist counselling were determined according to additional medical fees for physicians’ prescriptions and explanation of benefit and risk of anti-cancer drugs in Japan.

Table 1. Attributes and Levels of Oncology Pharmacy Counselling

Attitude of pharmacist
Unfriendly and uninterested
Friendly but distant
Friendly and interested
Quality of information
Unclear and contradictory information
Only general information
Clear and customized information
Explanation of side effects
Emphasis on life-threatening side effects
Emphasis on effects on normal life style, including cosmetic problems
Frequency of pharmacist counselling before starting chemotherapy
Once
More than once
Cost of pharmacist counselling
0 yen ($0)
250 yen (about $2.5)
500 yen (about $5)
Follow-up with the pharmacist after starting chemotherapy
At physician’s, nurse’s, or pharmacist’s request
At patient’s request
Every time a patient visits

Construction of the Questionnaire

The combination of attributes and levels in this study was 3⁴×2²=324 possible scenarios (in a full factorial design). To reduce the number of scenarios without sacrificing validity, a Bayesian nonlinear optimal design was employed to obtain 16 scenarios for the pilot study. These scenarios were divided into two questionnaires of eight choice sets by using JMP. Each DCE questionnaire included questions about marital status (married, unmarried, divorced, and bereaved), level of education (low [elementary: grades 1–6 and junior high school: grades 7–9], middle [high school: grades 10–12], and high [university or graduate school: grade 13 or above]), employment status (full-time, part-time, housekeeper, and unemployed or retired), household size (living with a partner, living alone, or other), and whether the patient had received pharmacist counselling in the past.

Pilot Study

Six patients who met the eligibility criteria completed one of the questionnaires, followed by structured interview and cognitive debriefing to identify uncertainty about question meanings, difficulty in answering questions, readability of questions, and the overall burden of completing the questionnaire. Additionally, we evaluated what patients thought about the attributes and levels and decided whether the attributes and levels should be retained or changed. The results of the pilot study were summarized and reviewed. The pilot study indicated that the attributes, levels, and wording were appropriate. However, responding to eight scenarios was revealed to be a burden to the patients. In consequence, we chose 12 scenarios on the basis of the patients’ recommendations. These 12 scenarios were divided into two sets of six choice sets using JMP. A sample scenario is shown in Table 2.

Table 2. An Example of One Scenario from a DCE Questionnaire about Patient Preferences for Pharmacist Counselling

When you receive information about chemotherapy for the first time, which pharmacist would you prefer to receive the information from?
	Pharmacist A	Pharmacist B
Attitude of pharmacist	Unfriendly and uninterested	Friendly but distant
Quality of information	Unclear and contradictory information	Only general information
Explanation of side effects	Emphasis on life-threatening side effects	Emphasis on effects on normal lifestyle, including cosmetic problems
Frequency of pharmacist counselling before starting chemotherapy	More than once	Once
Cost of pharmacist counselling	0 yen ($0)	500 yen (about $5)
Follow-up with the pharmacist after starting chemotherapy	Every time a patient visits	At physician’s, nurse’s, or pharmacist’s request
Which would you prefer?	☐	☐

Main Study

Setting and Participants

The main study was conducted in the outpatient chemotherapy departments of two centres, the Tokyo Medical University Hospital and the Nagumo Clinic. The study involved presentation of hypothetical scenarios in which participants would receive information about undergoing chemotherapy for the first time. Because we felt that these scenarios might create a psychological burden for participants diagnosed with breast cancer who had received no prior chemotherapy, we excluded these patients. The criteria for eligibility were as follows: female, 20 years of age and older, diagnosed with breast cancer, and outpatient who had previously received or was currently receiving chemotherapy. The exclusion criteria were as follows: Eastern Cooperative Oncology Group Performance Status 3 or 4, cognitive disorder, agitation or delirium, severe psychiatric disorder, or patients judged inappropriate (e.g. patients’ burden, not severe but problematic psychiatric concern) for this study by a health care provider. The sample size calculation was based on a measure of choice probability with some desired level of accuracy. By referring to Louviere et al.,²⁷⁾ it was calculated that 1537 scenarios would be needed to estimate a true proportion of 0.2 (the proportion in an Asian population of a frequent indicator of chemotherapy,²⁸⁾ within 10% of the true value 0.2 with a 95% confidence level.²⁷⁾ We initially planned to use two questionnaires of eight scenarios each, requiring a total of 192 patients. However, based on the result of the pilot study, we changed the count of scenarios from eight to six on each questionnaire. To assess the 1537 scenarios needed for the study, we re-calculated the sample size, and a total of 257 patients would be needed. Finally, we set a target number of 140 patients for each questionnaire (a total of 280 patients) to account for incomplete questionnaires.

Data Collection

We used block randomization to ensure equal numbers of participants for each questionnaire in each centre. In outpatient chemotherapy departments, pharmacists obtained informed consent and asked patients to complete one of the DCE paper-and pencil self-administered questionnaires. Clinical information including age, presence or absence of metastasis or recurrence of cancer, and current status of chemotherapy and endocrine therapy were obtained from patients’ medical records.

Analysis

In the current actual practice of pharmacist counselling in Japan, patients actually do not have the alternative to consult with pharmacists. Thus, we hypothesized that our choice model is consistent with independence of irrelevant alternatives, and the binary response data from the DCE questionnaire were analysed using the conditional logit model. We selected the levels that we assumed patients would prefer as references. Wald chi-squared values were calculated separately for attributes and levels. Utilities and regression coefficients representing the degree of preference for each level were also calculated. Greater Wald chi-squared values indicated greater preferences for attributes, and greater differences from references for levels, respectively. Odds ratios (ORs) were calculated by exponentiating the regression coefficients. Greater regression coefficient values and ORs indicated greater preferences for levels. Relative importance of attributes was calculated by dividing the utility range by the sum of all utility ranges for all attributes. Thus, an attribute with a wider range will have high relative importance. The utilities and relative importance of stratified data (chemotherapy, age, recurrence, marital status, previous pharmacist counselling) were also calculated. Descriptive statistics including mean and standard deviation were used to summarize patients’ demographics. All statistics were analysed with SAS (version 9.1.3, SAS Institute Inc.).

RESULTS

Participants

A total of 280 outpatients with breast cancer participated in this study between November 2012 and April 2013. Two patients were excluded from the analysis, one for consent withdrawal and one for lack of response. One patient did not answer questions about one scenario but was included in the analysis. Finally, the responses of 278 patients to 1667 scenarios were analysed.

Participants’ characteristics are shown in Table 3. The average (standard deviation) age was 53.0 (10.7) years with a range of 28–80 years. About half of the patients had regional recurrence (9.4%) or metastasis (39.6%) of their cancer. A total of 164 patients (59.0%) were undergoing chemotherapy and 121 patients (43.5%) were undergoing endocrine therapy. The level of education was high or middle for the majority of patients. With regard to employment status, 92 (33.1%) and 66 (23.7%) of patients were engaged in full-time and part-time work, respectively. Two-thirds of patients (66.2%) lived with a partner. A large majority of patients (87.7%) had already experienced pharmacist counselling.

Table 3. Patients’ Characteristics (N=278)

Age (years)	Mean	S.D.
	53.0	10.7
	n	%
Recurrence or metastasis
None	142	51.1
Regional recurrence	26	9.4
Metastasis	110	39.6
Chemotherapy
Undergoing	164	59.0
Previously experienced	114	41.0
Endocrine therapy
Undergoing	121	43.5
Not experienced	118	42.4
Previously experienced	39	14.0
Marital status
Married	177	63.7
Unmarried	47	16.9
Divorced	38	13.7
Bereaved	16	5.8
Education level
High	69	24.9
Middle	198	71.5
Low	6	2.2
Other	4	1.4
Employment status
Full-time	92	33.1
Part-time	66	23.7
Housekeeper	102	36.7
Unemployed or retired	18	6.5
Household
Living with a partner	184	66.2
Living alone	72	25.9
Other	22	7.9
Pharmacist counselling
Previously experienced	242	87.7
Not experienced	34	12.3

Attributes and Levels Predicting Preferences for Pharmacist Counselling

The Wald chi-squared values and relative importance of each attribute are shown in Tables 4 and 5. The four attributes of ATTITUDE (p<0.0001, relative importance 26.9%), INFORMATION (p<0.0001, relative importance 37.1%), COST (p<0.0026, relative importance 7.6%), and FOLLOW-UP (p<0.0001, relative importance 24.4%) were significant predictors of choice. The highest Wald chi-squared value of 157.66 and relative importance of 37.1% indicated that the attribute of INFORMATION was most important to participants. The Wald chi-squared value of 81.40 and relative importance of 26.9% for ATTITUDE was nearly equal to the value of 80.92 and relative importance of 24.4% for FOLLOW-UP. COST had a significant but relatively low Wald chi-squared value of 11.94 and relative importance of 7.6%. The regression coefficients, odds ratios, and Wald chi-squared values and utilities for each level are also shown in Tables 4 and 5, respectively.

Table 4. Conditional Logistic Regression Models of Patients’ Preferences for Pharmacist Counselling

	Coefficient	Odds ratio (95% CI)	Wald	p-Value
Attitude of pharmacist			81.40*	<.0001
Unfriendly and uninterested	−1.2272	0.29 (0.22–0.39)	70.51	<.0001
Friendly but distant	−0.7125	0.49 (0.37–0.64)	26.51	<.0001
Friendly and interested	Reference
Quality of information			157.66*	<.0001
Unclear and contradictory information	−1.6912	0.18 (0.14–0.25)	133.45	<.0001
Only general information	−1.0579	0.35 (0.26–0.46)	53.17	<.0001
Clear and customized information	Reference
Explanation of side effects			2.94*	0.0862
Emphasis on life-threatening side effects	−0.1559	0.86 (0.72–1.02)	2.94	0.0862
Emphasis on effects on normal life style, including cosmetic problems	Reference
Frequency of pharmacist counselling before starting chemotherapy			0.10*	0.7527
Once	−0.0305	0.97 (0.80–1.17)	0.10	0.7527
More than once	Reference
Cost of pharmacist counselling			11.94*	0.0026
500 yen (about $5)	−0.3440	0.71 (0.58–0.87)	11.29	0.0008
250 yen (about $2.5)	−0.2547	0.78 (0.64–0.93)	7.24	0.0071
0 yen ($0)	Reference
Follow-up with the pharmacist after starting chemotherapy			80.92*	<.0001
At physician’s, nurse’s, or pharmacist’s request	−0.0125	0.99 (0.82–1.19)	0.02	0.8944
At patient’s request	1.0998	3.00 (2.28–3.96)	61.00	<.0001
Every time a patient visits	Reference

* Wald chi-squared values were calculated separately for levels and attributes.

Table 5. Utilities and Relative Importance

	Overall		Chemotherapy				Age				Recurrence or metastasis				Marital status
			Undergoing		Previously experienced		50 and less		Over 50		None		Recurrence or metastasis		Married		Unmarried
	Utility	RI	Utility	RI	Utility	RI	Utility	RI	Utility	RI	Utility	RI	Utility	RI	Utility	RI	Utility	RI
Attitude of pharmacist
Unfriendly and uninterested	−0.58	26.9	−0.54	23.6	−0.62	30.2	−0.63	24.1	−0.56	29.2	−0.60	27.7	−0.57	23.3	−0.59	25.8	−0.58	27.8
Friendly but distant	−0.07		−0.13		0.01		−0.06		−0.06		−0.12		−0.02		−0.07		−0.07
Friendly and interested	0.65		0.68		0.63		0.69		0.63		0.72		0.58		0.66		0.65
Quality of information
Unclear and contradictory information	−0.77	37.1	−0.86	36.8	−0.69	36.1	−0.95	37.4	−0.65	35.6	−0.85	35.8	−0.72	35.2	−0.80	36.3	−0.76	38.0
Only general information	−0.14		−0.19		−0.10		−0.14		−0.14		−0.02		−0.29		−0.14		−0.16
Clear and customized information	0.92		1.05		0.79		1.09		0.80		0.86		1.01		0.95		0.92
Explanation of side effects
Emphasis on life-threatening side effects	−0.08	3.4	−0.16	6.0	0.01	0.3	−0.12	4.2	−0.05	2.4	0.04	1.8	−0.22	8.9	−0.15	6.3	0.04	1.9
Emphasis on effects on normal life style, including cosmetic problems	0.08		0.16		−0.01		0.12		0.05		−0.04		0.22		0.15		−0.04
Frequency of pharmacist counselling before starting chemotherapy
Once	−0.02	0.6	−0.05	2.1	0.04	1.7	−0.02	0.6	−0.02	1.2	−0.02	1.0	−0.02	0.7	−0.02	0.8	−0.01	0.3
Morethanonce	0.02		0.05		−0.04		0.02		0.02		0.02		0.02		0.02		0.01
Cost of pharmacist counselling
500 yen (about $5)	−0.14	7.6	−0.15	7.3	−0.16	8.3	−0.13	9.4	−0.15	6.5	−0.01	5.9	−0.30	11.9	−0.11	6.5	−0.22	9.2
250 yen (about $2.5)	−0.06		−0.08		−0.02		−0.19		0.03		−0.14		0.00		−0.10		0.04
0 yen ($0)	0.20		0.23		0.18		0.32		0.12		0.14		0.29		0.21		0.19
Follow-up with the pharmacist after starting chemotherapy
At physician’s, nurse’s, or pharmacist’s request	−0.37	24.4	−0.40	24.2	−0.33	23.4	−0.42	24.2	−0.37	25.2	−0.47	27.9	−0.29	19.9	−0.40	24.3	−0.33	22.7
At patient’s request	0.74		0.83		0.63		0.87		0.66		0.87		0.64		0.78		0.67
Every time a patient visits	−0.36		−0.42		−0.30		−0.45		−0.29		−0.40		−0.35		−0.38		−0.34

RI: relative importance.

Of the levels of ATTITUDE, patients preferred ‘friendly and interested’ over ‘friendly but distant’ (OR 0.49, 95% confidence interval (CI): 0.37–0.64), and ‘friendly but distant’ was preferred over ‘unfriendly and uninterested’ (OR 0.29, 95% CI: 0.22–0.39). ‘Clear and customized information’ was preferred over ‘only general information’ (OR 0.35, 95% CI: 0.26–0.46), and ‘only general information’ was preferred over ‘unclear and contradictory information’ (OR 0.18; 95% CI: 0.14–0.25) for INFORMATION. For the attributes of SIDE EFFECTS and FREQUENCY, the coefficients of each level were similar to the reference and the odds ratios were nearly 1.0. This means that there were no preferences between the levels for the attributes of SIDE EFFECTS and FREQUENCY. Coefficients for COST were both significant and the OR of the levels of ‘500 yen (about $5)’ (OR 0.71; 95% CI: 0.58–0.87) and ‘250 yen (about $2.50)’ (OR 0.78; 95% CI: 0.64–0.93) were nearly equal. The level of ‘0 yen’ was preferred over ‘500 yen’ and ‘250 yen.’ For FOLLOW-UP, the coefficient for ‘at patient’s request’ was positive, and the odds ratio for this level was 3.00 (95% CI: 2.28–3.96), indicating that patients would prefer follow-up at their request rather than ‘every time a patient visits’ or ‘at physician’s, nurse’s, or pharmacist’s requests.’ The differences of utilities and relative importance of patient’s characteristics are shown in Table 5. Regardless of these characteristics, INFORMATION was the most important attribute and the attributes of ATTITUDE and FOLLOW-UP were equally important after INFORMATION. The relative importance of the attribute of SIDE EFFECTS varied by patient characteristics; for example, patients who previously received chemotherapy attached less importance to SIDE EFFECTS (relative importance 0.3%) than FREQUENCY (1.7%), whereas patients currently undergoing chemotherapy placed more emphasis on SIDE EFFECTS (6.0%) than FREQUENCY (2.1%).

DISCUSSION

We studied preferences for pharmacist counselling among patients with breast cancer by using hypothetical scenarios in outpatient chemotherapy settings. Our results indicated that the attitude of the pharmacist, the quality of information provided by the pharmacist, the cost of the counselling, and the frequency of follow-up were significant predictors of patients’ preferences.

For the attribute of quality of information, the odds ratio for ‘unclear and contradictory information’ was very low (0.18). However, the odds ratio for ‘only general information’ was also low (0.35). This result indicates that patients tended to attach overriding importance to the quality of information provided by the pharmacist. By contrast, the attribute of ‘explanation of side effects’ was not a predictor of patient preferences. Patients tended to prefer the explanation of side effects with emphasis on normal life (OR 0.86, 95% CI: 0.72–1.02), but there was no statistically significant difference. Regarding the wording of the two levels of SIDE EFFECTS, the phrases ‘life-threatening side effects’ and the ‘influence on resuming normal lifestyle’ are plausible as suggested in the methodological review,²⁹⁾ and there were no problems pointed out by patients in the pilot study. However, two patients pointed out that age and a point of time during chemotherapy might influence their preference.

The importance of information for patients has been shown in research on information needs.^30–33) The results of a systematic review show that patients with fulfilled information needs have a better quality of life and less anxiety and depression.³⁴⁾ Degner et al. reported that newly diagnosed breast cancer patients place a high priority on information about chances of cure, spread of the disease, treatment options, family (genetic) risks, and adverse effects.³⁵⁾ These information needs vary with the type of cancer and stage of disease and can change during treatment.^36,37) Information needs that patients feel are unmet also change over time. Unmet information needs are relatively low in the diagnosis phase compared with that in the treatment phase.³⁸⁾ These facts indicate that our results, with no notable distinction in patients’ preferences between levels for the attribute of explanation of side effects, were possibly affected by recall bias. Furthermore, the levels of ‘emphasis on life-threatening side effects’ and ‘emphasis on effects on normal life style, including cosmetic problems’ are contradictory, but might be both important for patients. If these levels of SIDE EFFECTS made the choice difficult, patients could have ignored this attribute. The mind-set that both levels are important might have influenced the choices of patients. Furthermore, in regard to follow-up pharmacist visits, patients preferred these visits at their own request. This result differed from our thought that patients would prefer pharmacist counselling at each visit. Bakker et al. reported that patients experience two modes of communication, ‘doctor talk’ and ‘comprehensive’ style, when interacting with a health care provider.³⁹⁾ ‘Doctor talk’ occurs during routine check-ups and is usually initiated by the physician. Meanwhile, contacts initiated by the patients themselves, a preference indicated by the patients in our study, are conducted in ‘comprehensive’ mode, in which communication addresses not only physical but also psychosocial status. Patients who experience this ‘comprehensive’ mode in addition to ‘doctor talk’ express high satisfaction.³⁹⁾

Our results showed that breast cancer patients receiving chemotherapy prefer ‘comprehensive’ communication. The shift of chemotherapy to ambulatory care requires patients to take on greater responsibility for managing or monitoring themselves. Appropriate information should be provided not only at the time of diagnosis but also during treatment. Indeed, McKee et al. reported that 76% of patients in the United States requested visits to a pharmacist.⁵⁾ In inpatient settings in Japan, where the process of shared decision making remains unconventional, the team of physician, pharmacist, and health care professionals works cooperatively to communicate information about treatment to patients. In contrast, the role of pharmacists in outpatient settings is predominantly providing the right combination of anticancer agents rather than ambulatory care consultation. Considering this current state of daily practice in Japan, patients should have access to pharmacist counselling during chemotherapy in ambulatory care settings.

Further research about the information needs of Japanese cancer patients is needed. We found that the attitude of the pharmacist and follow-up visits with the pharmacist were of the same degree of importance to patients. Similarly, Kamei et al. reported that pharmacist attitude is strongly related to patient satisfaction.⁴⁰⁾ Pharmacist attitudes have been investigated in regard to pharmacist-patient communication,^40,41) but most of this research has been conducted in community pharmacy settings.⁴²⁾ In these settings, pharmacists tend to provide information about the name of the medication, usage instructions, and typical side effects without spending sufficient time assessing and communicating with patients.^43,44) Because information about chemotherapy should address side effects as well as psychological or social factors, it is necessary for pharmacists to understand the patient’s social background and perceptions. Greater attention to the psychological or social effects of medication and efforts to have a dialogue with patients may lead to relationship-building and beneficial changes in attitudes toward pharmacists.

Previous studies have shown that patients are willing to pay for pharmaceutical services.^45–47) Our odds ratio results showed that a cost of 500 yen would have only a small impact on patients’ preferences. Although there has been no fee for pharmacist counselling in oncology settings in Japan, the Ministry of Health, Labour and Welfare confirmed a fee for credentialed oncology pharmacists in the revision of medical treatment fees for 2014. The self-pay burden of the cost of explanation about chemotherapy to outpatients or inpatients by pharmacist is 600 yen (30% of 2000 yen) in this revision. In our study, the cost means the self-pay burden; thus, the result of COST hypothesized in our study is easily generalizable. Our results indicate that some patients in Japan are willing to pay for pharmacist counselling before initiating chemotherapy if pharmacists assume an engaged attitude and provide clear and customized information. Furthermore, patients attached a higher value to follow-up with proper timing after starting chemotherapy, namely management of chemotherapy-induced adverse events, than cost. These findings support the need for pharmacist counselling in oncology settings in Japan.

Our study has several limitations. First, the participants were recruited from only two centres; therefore, our results may not be generalizable. Second, the hypothetical scenarios used in this stated preferences study could cause recall bias. The experience of pharmacist or medical oncologist counselling can particularly affect patient’s choice. This probability cannot be excluded. Third, the attributes of ‘frequency of pharmacist counselling before starting chemotherapy’ and of ‘explanation of side effects’ were found to be inconsequential. Insufficient sample size in the pilot study was suspected of contributing to this problem of questionnaire design; thus, the design of the questionnaire may require further improvement. Fourth, we specifically designed to have good estimates for main effects of each attribute, and excluded the interactive effects in our model.

CONCLUSION

Our results indicate that patients receiving pharmacist counselling before starting chemotherapy prefer to interact with a pharmacist with a friendly, interested attitude who provides individualized information at no cost. Further research is needed to enhance pharmacist–patient communication.

Acknowledgment

We acknowledge all participating patients for their invaluable assistance. We also acknowledge and thank our collaborators for their thoughtful contribution to the conduct of this study (Hiroshi Kaise, Akiko Kaneko, Yuriko Sato, Atsuko Nakagawa, Aki Kitano, Akiko Takahashi, Chinatsu Iga, Hideaki Ayuhara, Nao Tanaka, and Ayako Mizoguchi).

REFERENCES

• 1) Matsuda A, Matsuda T, Shibata A, Katanoda K, Sobue T, Nishimoto H; Japan Cancer Surveillance Research Group. Cancer incidence and incidence rates in Japan in 2007: a study of 21 population-based cancer registries for the Monitoring of Cancer Incidence in Japan (MCIJ) Project. Jpn. J. Clin. Oncol., 43, 328–336 (2013).
• 2) Whitmer KM, Pruemer JM, Nahleh ZA, Jazieh AR. Symptom management needs of oncology outpatients. J. Palliat. Med., 9, 628–630 (2006).
• 3) Yamagishi A, Morita T, Miyashita M, Kimura F. Symptom prevalence and longitudinal follow-up in cancer outpatients receiving chemotherapy. J. Pain Symptom Manage., 37, 823–830 (2009).
• 4) Kawaguchi T, Azuma K, Yamaguchi T, Soeda H, Sekine Y, Koinuma M, Takeuchi H, Akashi T, Unezaki S. Development and validation of the Japanese version of the Decisional Conflict Scale of the value of pharmacists’ information: a before and after study. BMC Med. Inform. Decis. Mak., 13, 50 (2013).
• 5) McKee M, Frei BL, Garcia A, Fike D, Soefje SA. Impact of clinical pharmacy services on patients in an outpatient chemotherapy academic clinic. J. Oncol. Pharm. Pract., 17, 387–394 (2011).
• 6) Gourdji I, McVey L, Loiselle C. Patients’ satisfaction and importance ratings of quality in an outpatient oncology center. J. Nurs. Care Qual., 18, 43–55 (2003).
• 7) Rehnberg G, Absetz P, Aro AR. Women’s satisfaction with information at breast biopsy in breast cancer screening. Patient Educ. Couns., 42, 1–8 (2001).
• 8) Mallinger JB, Griggs JJ, Shields CG. Patient-centered care and breast cancer survivors’ satisfaction with information. Patient Educ. Couns., 57, 342–349 (2005).
• 9) Cox A, Jenkins V, Catt S, Langridge C, Fallowfield L. Information needs and experiences: an audit of UK cancer patients. Eur. J. Oncol. Nurs., 10, 263–272 (2006).
• 10) Davies NJ, Kinman G, Thomas RJ, Bailey T. Information satisfaction in breast and prostate cancer patients: implications for quality of life. Psychooncology, 17, 1048–1052 (2008).
• 11) Liekweg A, Westfeld M, Jaehde U. From oncology pharmacy to pharmaceutical care: new contributions to multidisciplinary cancer care. Support. Care Cancer, 12, 73–79 (2004).
• 12) Doloresco F, Vermeulen LC. Global survey of hospital pharmacy practice. Am. J. Health Syst. Pharm., 66 (Suppl. 3), S13–S19 (2009).
• 13) de Oliveira DR, Shoemaker SJ. Achieving patient centeredness in pharmacy practice: openness and the pharmacist’s natural attitude. J. Am. Pharm. Assoc. (2003), 46, 56–64, quiz, 64–66 (2006).
• 14) Iihara H, Ishihara M, Matsuura K, Kurahashi S, Takahashi T, Kawaguchi Y, Yoshida K, Itoh Y. Pharmacists contribute to the improved efficiency of medical practices in the outpatient cancer chemotherapy clinic. J. Eval. Clin. Pract., 18, 753–760 (2012).
• 15) Ikesue H, Oishi R. Oncology pharmacy specialists in oncology. Gan To Kagaku Ryoho, 35, 578–582 (2008).
• 16) Ben-Akiva M, Lerman SR. Discrete choice analysis: theory and application to travel demand. MIT Press, Cambridge, MA (1985).
• 17) Hensher DA. Stated preference analysis of travel choices: the state of practice. Transportation, 21, 107–133 (1994).
• 18) Propper C. Contingent valuation of time spent on NHS waiting lists. Econ. J., 100, 193–199 (1990).
• 19) Naik-Panvelkar P, Armour C, Saini B. Discrete choice experiments in pharmacy: a review of the literature. Int. J. Pharm. Pract., 21, 3–19 (2013).
• 20) Bridges JF, Hauber AB, Marshall D, Lloyd A, Prosser LA, Regier DA, Johnson FR, Mauskopf J. Conjoint analysis applications in health—a checklist: a report of the ISPOR Good Research Practices for Conjoint Analysis Task Force. Value Health, 14, 403–413 (2011).
• 21) Kessels R, Jones B, Goos P, Vandebroek M. The usefulness of Bayesian optimal designs for discrete choice experiments. Appl. Stochastic Models Bus. Ind., 27, 173–188 (2011).
• 22) Porteous T, Ryan M, Bond CM, Hannaford P. Preferences for self-care or professional advice for minor illness: a discrete choice experiment. Br. J. Gen. Pract., 56, 911–917 (2006).
• 23) Kimman ML, Dellaert BG, Boersma LJ, Lambin P, Dirksen CD. Follow-up after treatment for breast cancer: one strategy fits all? An investigation of patient preferences using a discrete choice experiment. Acta Oncol., 49, 328–337 (2010).
• 24) van Empel IW, Dancet EA, Koolman XH, Nelen WL, Stolk EA, Sermeus W, D'Hooghe TM, Kremer JA. Physicians underestimate the importance of patient-centredness to patients: a discrete choice experiment in fertility care. Hum. Reprod., 26, 584–593 (2011).
• 25) Papanikolaou P, Lyne P, Ratcliffe J. Using the discrete choice experimental design to investigate decision-making about pressure ulcer prevention by community nurses. Health Soc. Care Community, 15, 588–598 (2007).
• 26) Mühlbacher AC, Nübling M. Analysis of physicians’ perspectives versus patients’ preferences: direct assessment and discrete choice experiments in the therapy of multiple myeloma. Eur. J. Health Econ., 12, 193–203 (2011).
• 27) Louviere JJ, Hensher DA, Swait JD. Stated choice methods: analysis and applications. Cambridge University Press, New York (2000).
• 28) Chuang E, Christos C, Flam A, McCarville K, Forst M, Shin S, Vahdat L, Swistel A, Simmons R, Osborne M, Moore A, Mazumdar M, Klein P. Breast cancer subtypes in Asian–Americans differ according to Asian ethnic group. J. Immigr. Minor. Health, 14, 754–758 (2012).
• 29) Lancsar E, Louviere J. Conducting discrete choice experiments to inform healthcare decision making: a user’s guide. Pharmacoeconomics, 26, 661–677 (2008).
• 30) Cassileth BR, Volckmar D, Goodman RL. The effect of experience on radiation therapy patients’ desire for information. Int. J. Radiat. Oncol. Biol. Phys., 6, 493–496 (1980).
• 31) Iconomou G, Viha A, Koutra A, Vagenakis AG, Kalofonos HP. Information needs and awareness of diagnosis in patients with cancer receiving chemotherapy: a report from Greece. Palliat. Med., 16, 315–321 (2002).
• 32) Galloway S, Graydon J, Harrison D, Evans-Boyden B, Palmer-Wickham S, Burlein-Hall S, Rich-van der Bij L, West P, Blair A. Informational needs of women with a recent diagnosis of breast cancer: development and initial testing of a tool. J. Adv. Nurs., 25, 1175–1183 (1997).
• 33) Husson O, Thong MS, Mols F, Oerlemans S, Kaptein AA, van de Poll-Franse LV. Illness perceptions in cancer survivors: what is the role of information provision? Psychooncology, 22, 490–498 (2013).
• 34) Husson O, Mols F, van de Poll-Franse LV. The relation between information provision and health-related quality of life, anxiety and depression among cancer survivors: a systematic review. Ann. Oncol., 22, 761–772 (2011).
• 35) Degner LF, Kristjanson LJ, Bowman D, Sloan JA, Carriere KC, O’Neil J, Bilodeau B, Watson P, Mueller B. Information needs and decisional preferences in women with breast cancer. JAMA, 277, 1485–1492 (1997).
• 36) Degner LF, Davison BJ, Sloan JA, Mueller B. Development of a scale to measure information needs in cancer care. J. Nurs. Meas., 6, 137–153 (1998).
• 37) Vogel BA, Bengel J, Helmes AW. Information and decision making: patients’ needs and experiences in the course of breast cancer treatment. Patient Educ. Couns., 71, 79–85 (2008).
• 38) Harrison JD, Young JM, Price MA, Butow PN, Solomon MJ. What are the unmet supportive care needs of people with cancer? A systematic review. Support. Care Cancer, 17, 1117–1128 (2009).
• 39) Bakker DA, Fitch MI, Gray R, Reed E, Bennett J. Patient-health care provider communication during chemotherapy treatment: the perspectives of women with breast cancer. Patient Educ. Couns., 43, 61–71 (2001).
• 40) Kamei M, Teshima K, Fukushima N, Nakamura T. Investigation of patients’ demand for community pharmacies: relationship between pharmacy services and patient satisfaction. Yakugaku Zasshi, 121, 215–220 (2001).
• 41) Tinelli M, Bond C, Blenkinsopp A, Jaffray M, Watson M, Hannaford P, Community Pharmacy Medicines Management Evaluation Team. Patient evaluation of a community pharmacy medications management service. Ann. Pharmacother., 41, 1962–1970 (2007).
• 42) Murad MS, Chatterley T, Guirguis LM. A meta-narrative review of recorded patient–pharmacist interactions: exploring biomedical or patient-centered communication? Res. Social Adm. Pharm., 10, 1–20 (2014).
• 43) Cavaco AM, Romano J. Exploring pharmacists’ communication with customers through screening services. Patient Educ. Couns., 80, 377–383 (2010).
• 44) Sleath B. Pharmacist question-asking in New Mexico community pharmacies. Am. J. Pharm. Educ., 59, 374–376 (1995).
• 45) Barner JC, Mason HL, Murray MD. Assessment of asthma patients’ willingness to pay for and give time to an asthma self-management program. Clin. Ther., 21, 878–894 (1999).
• 46) Gore PR, Madhavan S. Consumers’ preference and willingness to pay for pharmacist counselling for non-prescription medicines. J. Clin. Pharm. Ther., 19, 17–25 (1994).
• 47) Suh DC. Consumers’ willingness to pay for pharmacy services that reduce risk of medication-related problems. J. Am. Pharm. Assoc. (Wash.), 40, 818–827 (2000).