Conflict of Interest: None
Development and Construct Validation of the Pharmacists' Care of Migraineurs Scale
Article first published online: 29 OCT 2008
© 2008 the Authors. Journal compilation © 2008 American Headache Society
Headache: The Journal of Head and Face Pain
Volume 49, Issue 1, pages 54–63, January 2009
How to Cite
Skomo, M. L., Desselle, S. P. and Shah, N. (2009), Development and Construct Validation of the Pharmacists' Care of Migraineurs Scale. Headache: The Journal of Head and Face Pain, 49: 54–63. doi: 10.1111/j.1526-4610.2008.01297.x
Financial support information: This study was funded by a Faculty Development Grant from Duquesne University.
- Issue published online: 5 JAN 2009
- Article first published online: 29 OCT 2008
- Accepted for publication September 14, 2008.
- migraineur care;
Objectives.— To develop the pharmacists' care of migraineurs scale (PCMS) and to evaluate its psychometric properties.
Background.— Migraine is often managed suboptimally in primary care. Migraineurs frequently come into contact with community pharmacists, who have the opportunity to make a positive impact on migraineur treatment outcomes. A valid and reliable tool that measures and documents the care provided by pharmacists to migraineurs is critical to the development and evaluation of educational programs and interventions.
Methods.— Relevant domains of pharmacist care and their respective composite items (behaviors) were identified through an extensive literature search and the use of 2 pharmacist and 2 migraineur focus groups sessions. The resultant 45 PCMS items composed a survey questionnaire mailed to a nationwide random sample of 6000 pharmacists. Data were subjected to an exploratory principal axis factoring procedure to discern the factor structure, and as such describe the latent domains composing the pharmacist caring behaviors constructs.
Results.— A total of 580 usable responses were returned, with an additional 60 returned as undeliverable, thus yielding a response rate of 9.7%. Exploratory factor analysis using principal axis factoring yielded 9 factors. However, upon examining the scree plot, communalities, and factor loadings, a reanalysis forcing a 7-factor solution yielded a more interpretable and plausible factor structure. The 7-factor solution included the following domains: (1) empathy; (2) prospective drug utilization review for newly diagnosed migraineurs; (3) medication counseling; (4) nonpharmacologic treatment plan; (5) headache sufferer triage; (6) dissemination of public health information; (7) maintenance of knowledge on migraine. Following the application of scale purification procedures, the final instrument is composed of 41 items and demonstrated a Cronbach's alpha reliability of 0.947. Cronbach's alpha reliabilities for the 7 domains ranged from 0.67 to 0.91, indicative of good to excellent internal consistency reliabilities for all the domains.
Conclusions.— The PCMS demonstrated very good construct validity and reliability. While additional validity testing is warranted, the PCMS should allow for benchmarking in the evaluation of interventions designed to improve pharmacists' care to migraineurs and for identifying correlates to effective community pharmacist migraineur care.
pharmacists' care of migraineurs scale
prospective drug utilization review
Migraine is a common, chronic, disabling condition characterized by recurrent episodes of headache sometimes accompanied by nausea, vomiting, photophobia, and phonophobia.1 It has been estimated that 17.1% of women and 5.6% of men in the United States suffer from migraines.2 Migraines have a negative effect on all aspects of quality of life including the physical, emotional, and social aspects.3 The Genetic Epidemiology of Migraine study found that migraineurs had significantly lower scores in the physical functioning, physical role limitations, mental health, pain, vitality, and general health perceptions dimensions compared with controls.4 Additionally, approximately two-thirds of migraineurs report some sort of disability associated with their headaches.5,6
Migraine is often managed suboptimally in primary care, with only about 50% of migraineurs seeking physician consultation.7,8 Many patients do not seek physician consultation for migraine because they believe they are suffering from other types of headaches or they doubt their physician can do anything to improve their headaches.7 Other barriers to effective care of migraineurs include incorrect diagnosis and suboptimal treatment regimens when the correct diagnosis has been made.7,9,10,11
Although migraine sufferers formally seek consultation by appointment with physicians, they also come into frequent contact with other healthcare professionals, such as pharmacists.12 In the United States, pharmacists make over 14 million over-the-counter (OTC) headache product recommendations in community pharmacies each year.13 This statistic illustrates the numerous instances where pharmacists are called upon to triage potential and actual migraine sufferers. Considering that only 19% of migraineurs report their migraine therapy to be effective,14 and an overwhelming 57% use only OTC medications to treat their condition,9,14-16 a tremendous opportunity rests with pharmacists to make a positive impact on the lives of migraineurs.
A thorough search uncovered only a paucity of literature evaluating pharmacists' understanding of and attitudes toward migraine and failed to identify constructs used to assess pharmacist care of the migraineur. One study used an instrument to evaluate pharmacists' knowledge, attitudes, and practice patterns concerning migraine therapy.17 This project concluded that while community pharmacists are well-positioned to improve migraineur care, many of them may not be well versed in the disease itself or with current migraine treatment guidelines.17 These results point to the need to develop migraine educational and clinical interventions aimed to improve migraineur care and the health outcomes of migraineurs. A key to success in designing and measuring the outcomes of such interventions is the use of a valid and reliable tool that measures and documents the level of care provided by pharmacists. Such a tool might be used for benchmarking and with other information to determine factors associated with high/low quality of care and improvements in care provided to recurrent headache sufferers. The objectives of this study were to (1) develop the pharmacists' care of migraineurs scale (PCMS); and (2) evaluate its psychometric properties.
This study was approved by the Institutional Review Board at the Duquesne University.
Identifying Relevant Domains of Migraineur Care by Pharmacists.— First, it was necessary to identify putative, relevant domains of pharmacists care.18 An extensive review of literature was conducted. The literature search was conducted on International Pharmaceutical Abstracts, Medline (PubMed), CINAHL, ERIC, PsycINFO, and Social Sciences Index databases employing the terms “migraine,”“migraine epidemiology,”“empathy”[caring, compassion]; “disease management”[therapy], “questionnaire”[questionnaire design], “pharmacist,”“pharmacy”[community pharmacy services], drug utilization review [drug utilization evaluation, drug-use review], triage, patient advocacy, and pharmacist knowledge (of) migraine. Information was sought from editorials, commentaries, and review articles in addition to reports of empirical studies. The search was expanded to include other articles not found in the initial review but referenced within articles uncovered in the initial review. There were no time (date of publication) restrictions placed on the search. Having relatively sparse information on pharmacist migraineur care in the existing literature, the investigators turned to a few key resources to assist in identifying possible domains of pharmacists' care of migraine sufferers.17,19-26 These studies on empathy, drug utilization review, and pharmacists' roles in medication therapy management were informative in proffering initial domains comprising the pharmacists' care of migraineurs survey instrument. The initially hypothesized domains were:
- 2Prospective drug utilization review (PDUR) for newly diagnosed migraineurs.
- 3Medication counseling for newly diagnosed migraineurs.
- 4Nonpharmacologic treatment plan.
- 5Monitoring of drug therapy effectiveness for “refill” or “return” migraineurs.
- 6Advocacy for migraineurs (eg, taking action when therapy is inadequate).
- 7Headache sufferer triage (eg, screening and recommendation to visit physician if headache is potentially associated with migraine).
- 8Public health information (eg, educational materials/seminars about migraine available to the public).
- 9Maintenance of knowledge on migraine (eg, staying abreast of current developments in migraine and migraine therapy).
Generation of Items Comprising Each Domain and Development of Survey Instrument.— A pool of items for each domain was developed using a careful and thorough review of the literature in addition to information acquired from the use of the groups. As recommended by Briggs et al,27 focus group sessions were conducted eliciting the opinions of 2 unique populations with potentially much to offer toward informing the phenomenon of interest (in this case, migraineur care): (1) pharmacists, who are in a ready position to describe the types of care that they provide to migraine sufferers and ultimately would like to provide in ideal situations; (2) migraineurs, who will be able to identify caring behaviors of pharmacists that have potentially affected their migraine therapy treatment outcomes.28 Two pharmacist and 2 migraineur focus groups were conducted. Based on the information gleaned from the 4 focus group sessions and the literature review, an initial PCMS measure composed of 43 items was developed. This initial set of items was pilot tested on a convenience sample of 40 community pharmacists. Based upon pilot testing, 2 items were added and other items were edited for clarity. The resultant, modified set of 45 items was purported to comprise the PCMS. Each item represents a pharmacists' migraineur care behavior and is measured using a scale where 1 = never to very rarely, 2 = rarely, 3 = sometimes, 4 = fairly often, and 5 = nearly all the time to all the time.
Mail Survey Questionnaire Administration.— A questionnaire composed of the PCMS items, questions eliciting demographic data, and items composing other variables was mailed to a nationwide random sample of 4000 pharmacists. A list of 10,000 community pharmacists was acquired from a commercial vendor (KM Lists) from which 4000 random numbers were generated using the website http://www.randomizer.org To maximize survey response rate, a modified Dillman29 approach of repeated contacts was used. The pharmacists sampled were first mailed a survey kit, which included the questionnaire along with a cover letter requesting participation and a self-addressed postage paid return envelope. Each envelope was coded by an administrative assistant with a number corresponding to a pharmacist to maintain anonymity of the participants. Nonresponders identified by the number code on the envelope were sent a reminder postcard at the end of the first 3 weeks since the mailing of the initial survey kit. Further nonresponders were sent an additional survey kit at the end of 2 weeks of mailing the reminder postcards with another cover letter. After an initial low rate of return, an additional 2000 pharmacists were mailed the survey using the same modified Dillman29 approach 5 months after the initial surveys were sent out so as to generate enough responses to conduct the appropriate statistical tests. Responses were input in spss® 13.0 for analysis.
Data Analysis.— Data were subjected to an exploratory principal axis factoring procedure to discern the factor structure, and as such describe the latent domains composing the pharmacist caring behaviors constructs. While the investigators hypothesized the domains composing the PCMS, it was believed that creation of a new measure necessitated an exploratory, rather than confirmatory approach in the analysis. For the procedure, an orthogonal (varimax) rotation was applied to allow the factors to correlate and assist with interpretation of the data, in accordance to an a priori expectation that the factors, all representing various aspects of care, would be highly related. The Kaiser criterion, which suggests including all factors with an eigenvalue greater than 1 in the final model, and an examination of a scree plot, which seeks to “identify the last substantial drop in the magnitude of eigenvalues,” were used to discern the optimal number of factors.30 A value of 0.40 was established as the cut-off point for significant factor loading, even though it has been argued that loadings as low as 0.32 are appropriate.31
The measure was further examined and purified through the use of procedures recommended by Nunnally, including the calculation of item-to-total correlations, the Cronbach's alpha for each domain, and the resulting Cronbach's alpha pending removal of each item from the domain, to assess internal consistency.31
Respondent Characteristics.— A total of 580 usable responses were returned, with an additional 60 returned as undeliverable, thus yielding a response rate of 9.7%. The respondents were mostly male (61.7%), and the mean age of the participants was 47.5 years. Respondents reported practicing as a community pharmacist for an average of 22.2 years. A plurality (192, 33.4%) of the respondents reported practicing in independent community pharmacy, followed by a large number working in large chain pharmacies (185, 31.9%). Demographic characteristics of the respondents are shown in Table 1.
|Characteristics||N = 580|
|Age (years), mean ± SD||47.53 ± 11.90|
|Gender, n (%)|
|Primary practice setting, n (%)|
|Large chain||185 (31.9)|
|Supermarket pharmacy (eg, Kroger, Albertsons)||82 (14.1)|
|Mass merchandiser (eg, Target, Wal-Mart)||71 (12.5)|
|Small chain||23 (3.9)|
|Number of years as community pharmacist, mean ± SD||22.21 ± 12.49|
|Number of years with current employer, mean ± SD||12.94 ± 10.16|
The potential for nonresponse bias was assessed through a wave analysis, whereby characteristics and responses of the first 20% of responders (early) and the last 20% responders (late) are compared, as late responders are said to resemble, or mimic nonresponders.32 Chi-square analysis of gender showed that there were no differences between the early and late responders (P = .170). The mean age of the early responders (48.34 ± 12.41) was not significantly different from the mean age of the late responders (47.09 ± 12.08 years). The mean PCMS score of the early responders (128.82 ± 23.89) was not statistically different (P = .482) from the mean PMCS score of the late responders (131.24 ± 24.61).
Exploratory Factor Analysis.— Exploratory factor analysis using principal axis factoring yielded 9 factors. However, upon examining the scree plot (Fig.), communalities, and factor loadings, a reanalysis forcing a 7-factor solution yielded a more interpretable and plausible factor structure, which differed only somewhat from the structure originally hypothesized. The amount of variance explained by the 7-factor solution (65.08%) was nearly as high as the variance explained by the 9-factor solution (66.22%), and the Cronbach's alpha reliability values for each of the domains improved with the 7-factor solution. Three poorly performing items were removed because of low factor loadings or significant cross-loading on more than one factor. These items concerned explaining to migraineurs the importance of eliciting social support, documenting migraineurs' medical information electronically, and providing written counseling information. One additional item dealing with conducting public seminars on migraine was removed from the scale, as it decreased the variance explained by the entire scale as well as the reliability of the factor to which it loaded. The 7-factor solution included the following domains:
- 2Prospective drug utilization review for newly diagnosed migraineurs.
- 3Medication counseling.
- 4Nonpharmacologic treatment plan.
- 5Headache sufferer triage.
- 6Dissemination of public health information.
- 7Maintenance of knowledge on migraine.
Table 2 provides the factor loadings of the final 41 items onto the 7 factors. Cronbach's alpha reliability for the overall measure was 0.947. Cronbach's alpha reliabilities for the 7 domains ranged from 0.67 to 0.91, indicative of good to excellent internal consistency reliabilities for all the domains.
|Factor (domain)||Factor loading||Mean‡|
|Factor 1: empathy (α = 0.844, overall mean = 3.90)†|
|I listen attentively to patients describe frustrations regarding their headaches.||0.680||3.91|
|I provide empathic responses to persons complaining of migraine or other recurring headaches.||0.752||4.10|
|I establish rapport with patients suffering from recurring headaches.||0.612||3.74|
|Factor 2: prospective drug utilization review (α = 0.759, overall mean = 4.19)†|
|I contact the prescriber if patients' prescribed therapy is not covered by insurance or whose cost may contribute to noncompliance.||0.497||4.26|
|I screen for drug-drug interactions before dispensing a drug used in migraine therapy.||0.751||4.49|
|I screen new migraine therapy prescriptions for potential drug-related problems.||0.710||4.28|
|I identify any adverse effects or other drug-related problems associated with patients' migraine therapy.||0.538||3.72|
|I contact the physician and/or health plan for prior authorization if the migraine sufferer needs more than the maximum amount of medication as covered by the health plan.||0.533||4.23|
|I contact the prescriber to discuss alternatives or prophylaxis when acute migraine therapy is insufficient or ineffective (eg, triptan not working, patient uses triptan frequently, migraines particularly severe and long duration).||0.544||2.85|
|Factor 3: headache sufferer triage (α = 0.808, overall mean = 2.50)†|
|I enlist the support of family or caregivers for patients suffering from migraine or recurring headaches.||0.505||2.25|
|I gather histories of consulting patients' recurring headache problems who have not yet seen a physician for their headache and make referrals if necessary.||0.703||2.22|
|I attempt to determine the nature of consulting patients' headaches to determine whether it is migraine or some other type of headache.||0.639||2.86|
|I refer patients with diagnosed but unmanaged headache problems to a specialist (eg, neurologist).||0.466||2.92|
|I instruct staff to alert me of someone who complains about or requests a remedy for recurring headaches.||0.666||2.43|
|I conduct a medication history review for newly diagnosed migraine sufferers.||0.510||2.99|
|Factor 4: medication counseling (α = 0.911, overall mean = 3.70)†|
|When dispensing a migraine patient's first triptan, I provide oral counseling on the proper directions for use of the medication, including the maximum dose per day and maximum dose per week.||0.786||4.10|
|When dispensing a migraine patient's first triptan, I provide oral counseling on headache rebound.||0.699||3.39|
|When dispensing a migraine patient's first triptan, I provide oral counseling on the possible adverse effects of the medication.||0.820||3.97|
|When a patient initially receives a prescription for migraine prophylaxis, I provide oral counseling on the importance of adherence to the regimen (taking the medication every day).||0.800||3.90|
|I inform patients receiving their first prescription for migraine prophylaxis that they may not notice results right away.||0.772||3.75|
|I inform patients receiving a new prescription for migraine prophylaxis that the medication may not completely eliminate their migraines.||0.775||3.72|
|I discuss potential effectiveness of prescribed therapies with migraine patients.||0.462||3.28|
|I look for signs of nonadherence to migraine prophylaxis (eg, late refills, increased migraine frequency and severity) among patients.||0.448||3.20|
|Factor 5: nonpharmacologic treatment plan (α = 0.912, overall mean = 2.90)†|
|I recommend to migraine sufferers that they use a headache diary to identify migraine patterns (frequency, severity, duration) and triggers.||0.566||2.47|
|I discuss identification and avoidance of common migraine triggers with migraine patients.||0.670||3.00|
|I discuss the role of OTC medications in the management of migraine with patients.||0.676||3.20|
|I discuss the role of complementary and alternative medicines in the management of migraine with patients.||0.708||2.41|
|I discuss nonpharmacologic adjunctive therapies for migraine attacks (heat, cold, relaxation techniques, rest, etc) with migraine patients.||0.699||2.79|
|I inquire about the effectiveness of OTC, alternative and complementary medicine, and other adjunctive therapy in ameliorating migraine attacks for patients.||0.693||3.06|
|I look for signs that migraine therapy is not as effective as it could be (eg, frequent use of OTC or prescription analgesics and triptans) for migraine patients.||0.563||3.14|
|I discuss alternatives with patients whose migraine therapy is ineffective.||0.608||3.34|
|I refer patients to credible educational information about migraine on the internet.||0.549||2.51|
|Factor 6: dissemination of public health information (α = 0.794, overall mean = 2.3)†|
|I provide written disease management information on migraine when dispensing prescriptions for migraine.||0.589||2.69|
|I make available brochures/pamphlets or other educational material about migraine in kiosks or other areas free for public consumption.||0.845||2.46|
|I refer patients to educational material available at the pharmacy.||0.805||2.50|
|I participate in screenings and/or migraine awareness programs.||0.522||1.55|
|Factor 7: maintenance of knowledge on migraine (α = 0.670, overall mean = 3.09)†|
|I attend live continuing education programming on migraine headache.||0.480||2.91|
|I complete self-study continuing education on migraine headache.||0.548||3.83|
|I read original research or review articles on migraine headache in peer-reviewed journals.||0.562||2.77|
|I seek information from drug sales representatives or medical/clinical liaisons on migraine headache.||0.662||3.03|
|I seek information on the internet regarding migraine headache.||0.625||2.90|
Scale Characteristics.— On a scale with a possible range of scores from 41 to 205, the mean PCMS score among respondents was 129.88 (±24.91), and the median of the scores was 129.00, which is only slightly more than the scale's median of 123. The lowest respondent score was 48, and the highest was 200. A closer examination of the distribution of PMCS scores revealed a slight, but nonsignificant positive skewness of 0.18, with a standard error of skewness equal to 0.11. Similarly, the PCMS scores revealed a positive, but very slight (nonsignificant) kurtosis of 0.13, and a standard error of kurtosis equal to 0.21. As such, the responses to the scale were normally distributed.33
Pharmacists have been successful in the management of various chronic diseases including diabetes,34,35 hypertension,36 and arthritis.37 While migraine is chronic in nature, few have looked at pharmacists' involvement specifically with recurrent headache sufferers. Measuring the care that community pharmacists provide to migraine sufferers provides important information regarding the impact community pharmacists have in the area of migraine. Psychometrically sound instruments that include relevant multi-item domains are necessary to assess pharmacist-provided care to migraineurs. The majority of migraine sufferers are not satisfied with their current therapy.14 Additionally, migraine patients who have not even been formally or properly diagnosed may come into frequent contact with pharmacists.13 These trends suggest that pharmacists may be in position to identify many more undiagnosed and under-treated migraine sufferers and also have the opportunity to uncover misconceptions about effective migraine therapy, both when filling prescriptions and talking with patients in the process of OTC medication purchases.
The domains and items of the PCMS were identified and developed through a comprehensive process, which included a comprehensive review of the migraine literature, migraine patient and community pharmacist focus groups, and input from a clinical expert and survey psychometrician. The 580 responses that were received were enough to conduct the factor analysis and minimized the sampling error, as it was greater than 10 times the number of items (45) in the original PCMS.31
The factor loadings of the 41 items reveal sufficient construct validity. There were no appreciable cross-loadings of the items that were selected within a factor with those of other items providing sufficient proof of the same. The reliability estimate obtained in any particular study is independent of the number of persons in the study, but in any study the reliability is directly related to the number of items on the test.31 The high Cronbach's alpha reliability values for the entire PCMS and its domains in this study provide sufficient evidence of the instrument's internal consistency reliability.
The PCMS should prove useful in evaluating the care provided by pharmacists to recurrent headache sufferers, particularly before and after educational, or other types of interventions. The PCMS might also be adapted for use in regard to other diseases, as there are common elements to pharmacists' care of patients, regardless of patient condition. Subscales of the PCMS might also be useful for more limited projects, for example, measuring empathy or knowledge. Subscales may be used simply by totaling respondents' scores of the items composing the domain of interest; however, a more accurate approach of discerning pharmacists' level of care on that domain would be to weight the responses to each item by that item's factor score (Table 2) acquired from running the principal axis factoring procedure.33
Each of the established domains of the PCMS is an important component of the overall care that pharmacists provide to migraine sufferers. Empathy is extremely important in the patient–healthcare provider relationship.21,22 Listening to patients complain about their headaches and providing empathic responses can help build relationships with patients, which can lead to better treatment outcomes. It has been demonstrated that pharmacists can learn empathic behaviors by observing and imitating others and consistently using self-reflective practices.38 Therefore, pharmacists who struggle with empathy can learn how to be empathic.
Prospective drug utilization review is an essential pharmacist function required by federal and most states' regulations. Pharmacists in this study generally reported to be actively engaged in this activity; however, given its legal, ethical, and patient outcome ramifications, it might be argued that even more frequent engagement is warranted.
Pharmacists have opined that in several instances they are the first healthcare providers patients consult.39 Therefore, pharmacists are in an ideal position to provide triage for chronic headache sufferers. One tool that can help pharmacists in triaging chronic headache sufferers is the use of migraine screening instruments. Hersberger et al have studied the screening function of community pharmacists for patients with sleep disorders.40 The goal of their study was to assist pharmacy clients in detecting likely causes of any sleep disturbance or daytime sleepiness through a free screening, and to deliver targeted counseling. Pharmacists were able to assess the pattern of individual sleep disorders and to identify a possible cause in nearly one-third of the cases. Migraine screening tools such as the ID Migraine41 would be quick and easy to administer at community pharmacists and would provide pharmacists with information necessary to effectively triage many chronic headache sufferers.
Many migraineurs can be successfully managed through appropriate lifestyle modifications, nonpharmacologic therapy, and pharmacologic therapy. Pharmacists are in an excellent position to provide not only medication counseling but information regarding migraine triggers, lifestyle modifications, and nonpharmacologic therapy. In general, patients with chronic diseases are not adherent to their medication regimens.42 Additional counseling and information provided by pharmacists regarding proper use of both acute and preventative migraine therapies could substantially increase migraineur treatment success rates.
Pharmacists have an opportunity to refer patients to educational material available in the pharmacy while they are filling the prescription, and also to make such material available. The utility and effectiveness of these materials on patient knowledge, treatment outcomes, and pharmacy repeat patronage merits further study.
In general, pharmacists indicated relatively high frequencies in PDUR and medication counseling behaviors but lower frequencies in nonpharmacologic treatment plan dissemination of public health information, and maintenance of knowledge on migraine behaviors. This might be explained by the legal requirements in the former, as well as those activities being historically thought of as “traditional” pharmacy services. The latter behaviors, however, are important to the effectiveness in care provided to migraine patients and might be considered key if pharmacy is to embrace a greater role in public health. Future study might examine reasons for pharmacists reporting lower frequencies of performing these behaviors and identify strategies by which these behaviors can be facilitated during practice.
Several limitations of the study must be considered. The generation of items for the PCMS was dependent on the focus group interviews and review of relevant literature. Alternative wording of items may have yielded different results. Limitations inherent to focus groups apply as well. One of these limitations is the potential presence of a moderator effect, wherein the participants are simply providing answers that seem desirable to the moderator. Another is a group effect, wherein the participant for fear of being singled out agrees with the opinions of the larger group, even though they may have an opposing idea.43 This study employed a purposive sampling procedure for recruiting both migraineurs and pharmacists to participate in the focus group sessions. However, a more diverse sample (ie, persons not affiliated with one university and more varied in age and race/ethnicity) might have provided richer commentary. The survey was initially mailed out to a nationwide sample of 4000 pharmacists and then an additional 2000. The response rates were similar; however, the generalizability of the results to a nationwide population of pharmacists is limited given the survey's relatively low rate of return. The main concern in regard to a low response rate is the potential for nonresponse bias, particularly with regard to disproportionately higher or lower responses by pharmacist position, years of experience, race/ethnicity, and other factors that may have influenced questionnaire recipients' practice experiences and attitudes. That being said, nonresponse bias could persist with response rates of up to 60%, or even higher.44 Additionally, the wave analysis confirmed that there were no differences between the early and the late responders to the survey. Still, the low rate of return may be a reflection of a lack of enthusiasm or support for the subject matter, perhaps indicative that nonresponders are potentially even less motivated and could conceptualize care of migraineurs differently. Another factor potentially contributing to the low rate of return was response burden in that the PCMS and its items were part of a larger survey of 106 items which may also have affected the rate of return. Finally, it should be noted that while the PCMS performed well in initial construct validity and reliability testing, further testing is needed to confer greater confidence in the convergent, discriminant, and criterion-related validity of the overall measure and each of its respective factors.
In conclusion, the creation of a valid and reliable tool to measure pharmacists' care of migraineurs can allow for benchmarking in the evaluation of interventions designed to improve pharmacists' care to migraineurs. Identifying correlates to effective community pharmacist migraineur care in the future will also assist in doing the same. Moreover, the identification of various demographic and practice environment variables related to pharmacists care of migraineurs can enable pharmaceutical companies and researchers to devise and target educational strategies to pharmacists (eg, education stressing empathy and/or knowledge) and also provide colleges and schools of pharmacy valuable information for instruction to professional pharmacy students in caring for migraineurs. This can manifest in substantial improvement in migraine-related health outcomes and patient quality of life.
- 1Headache Classification. Subcommittee of the International Headache Society. The International Classification of Headache Disorders, 2nd edition. Cephalalgia. 2004;24(Suppl. 1):1-151.
- 5Impact of migraine and tension-type headache on lifestyle, consulting behaviour, and medication use: A Canadian population survey. Can J Neurol Sci. 1998;20:131-137., , , , , .
- 11The role of concomitant headache types and non-headache co-morbidities in the underdiagnosis of migraine. Neurology. 2002;58(Suppl. 6):S3-S9..
- 13OTC products: 2004 survey of pharmacist recommendations. Pharmacy Times. 2004;(Suppl.):1-44.
- 19Taxonomy project moves pharmacy practice closer to a uniform language: PPAC system will facilitate research and provide a stronger foundation for billing systems. J Am Pharm Assoc. 1997;Nov-Dec:629-631., .
- 24Summary of the final report of the Scope of Pharmacy Practice project. Am J Hosp Pharm. 1994;51:2179-2182.
- 28Migraineurs' perceptions of the role of pharmacists in providing care. International Journal of Pharmacy Practice. 2008;16:1-7., , .
- 29Mail and Telephone Surveys: The Total Design Method. New York, NY: John Wiley & Sons; 1978..
- 30Using Multivariate Statistics. Boston, MA: Allyn and Bacon; 2001., .
- 31Psychometric Theory. New York: McGraw-Hill; 1994., .
- 33Statistical Methods for Health Care Research, 2nd edn. Philadelphia, PA: J.B. Lippincott Company; 1993., .
- 34Outcomes of a community pharmacy-based disbetes monitoring program. J Am Pharm Assoc. 1999;30:238-243., , , , , .
- 35The Asheville project: Short-term outcomes of a community pharmacy diabetes care program. J Am Pharm Assoc. 2003;43:149-159., .
- 39History taking enhances pharmacist ability to triage. ASHP Annual Meeting. 1998;55:22..
- 44Foundations of Behavioral Research, 4th edn. Orlando, FL: Harcourt College Publishers; 2000., .