American Journal of Therapeutics 10, 93–103 (2003)
Provision of Drug Information to Patients by Pharmacists:
The Impact of the Omnibus Budget Reconciliation Act of
1990 a Decade Later
Robin Schatz, Robert J. Belloto, Jr., Donald B. White, and Kenneth Bachmann*
Drug-related illness in the United States factors substantially in health care costs, although often
these illnesses and their attendant costs are preventable. One strategy for minimizing adverse drug
reactions is to provide drug information to consumers in the form of prescription counseling at
pharmacies. The Omnibus Budget Reconciliation Act of 1990 (OBRA 1990) contained provisions for
mandating such counseling to Medicaid patients. OBRA 1990 was implemented in 1993, but most
states acted quickly to extend counseling services to all patients receiving prescription drugs.
We looked at the extent and quality of prescription counseling available in community pharmacies
1 decade after OBRA 1990 was written. We evaluated the counseling services afforded at large chain
pharmacies, independent community pharmacies, and on-line pharmacies for a hydrochlorothiazide prescription. We found that most (69%) pharmacies offered to provide prescription counseling
service, and that average counseling index scores, a measure of the quality or extent of information
provided as determined by a Rasch analysis, were generally satisfactory.
Our observations based on a single prescription for hydrochlorothiazide, along with other studies,
suggest that there is a positive upward trend in the number of pharmacies providing prescription
drug information, and that the extent of information provided suggests that the objectives of OBRA
1990 and related legislation to reduce ADRs are being fundamentally satisfied.
Keywords: OBRA 1990, drug information, patient counseling, pharmacists.
INTRODUCTION
Drug-related illness in the United States is considered
to be a substantial factor in health care costs. In 1995,
drug-related illness was responsible for an estimated
17 million emergency room visits, 8.7 million hospital
admissions, and $76.6 billion in hospital costs.1 The
US Food and Drug Administration (FDA) reported
12,000 deaths caused by adverse drug reactions
(ADRs) in 1987.2
Department of Pharmacology, College of Pharmacy, The University of Toledo, Toledo, Ohio, USA.
*Address for correspondence: Department of Pharmacology, College of Pharmacy, The University of Toledo, 2801 W. Bancroft St.,
Toledo, OH 43606, USA. E-mail: kbachma@utnet.utoledo.edu
1075–2765 © 2003 Lippincott Williams & Wilkins, Inc.
Over one-fourth of the hospitalizations of the elderly have been ascribed to either ADRs or medication
noncompliance.3 Many of these ADRs and their sequelae of emergency room visits, hospitalizations, and
attendant costs are preventable, considering that in
approximately one-third of emergency room visits
precipitated by ADRs, patients had little or no understanding of how to properly self-administer their
medications.4 Lakshmanan et al reported that 5% of
admissions to an acute care facility were iatrogenic,
and 48% of those were preventable.5 The Adverse
Drug Event (ADE) Prevention Study Group demonstrated that nearly 30% of ADEs in a hospital setting
are preventable, especially if intercepted early in the
process.6 With regard to ambulatory elderly patients,
it has been estimated that potentially inappropriate
medications are prescribed for nearly one-fourth of
this cohort,7 and that at least one potentially inappropriate medication is prescribed in 4.5% of outpatient
94
visits by the elderly.8 In an analysis of 414 communitydwelling subjects aged 75 years and older in Santa
Monica, California who were not terminally ill, and
had no severe cognitive or functional impairments,
14% were found to be using at least one inappropriate drug.9
In an effort to reverse the escalating costs of drugrelated illness, Congress included language in the
Omnibus Budget Reconciliation Act of 1990 (OBRA
1990)10 that required pharmacists to offer to provide
drug information about their prescriptions (ie, to
counsel) to Medicaid patients. These requirements in
OBRA 1990 became effective on January 1, 1993. most
states quickly followed suit, but also required pharmacists to offer to counsel all patients, not just Medicaid patients. Ultimately, the obligation of pharmacists to counsel patients was included in the American
Pharmaceutical Association (APhA) Standard of Practice for the Profession of Pharmacy.11
As recently as 1997, an effort was made to assess the
impact of OBRA 1990 on the counseling practices of
community pharmacists in Nebraska.12 Almost half
(45%) of the pharmacy respondents reported that time
devoted to patient counseling had increased; however,
statutory obligations and practice standards to offer
counseling services have been in place for over 7 years
now. In view of the potential for effective pharmaceutical counseling to preclude or mitigate ADRs, this
study was undertaken with two objectives:
• To evaluate the extent to which pharmacists are
meeting their mandate to offer prescription counseling to patients;
• To determine whether practice setting (ie, independent community pharmacy, chain pharmacy,
or on-line pharmacy) is a factor in the quality or
quantity of prescription counseling services provided by pharmacists.
METHODS
Investigational Review Board
We submitted an application for review to the Human
Subjects Research Review Committee (HSRC) of the
University of Toledo. The HSRC reviewed the application and determined that this project was exempted
from HSRC review in accordance with Code 45CFR
46.101 of the Federal Regulations.
Pharmacies
We selected a total of 90 pharmacies for study as follows: 30 were independent community pharmacies, 30
American Journal of Therapeutics (2003) 10(2)
SCHATZ ET AL
were large chain pharmacies, and 30 were on-line
pharmacies. We selected on-line pharmacies from a
search engine’s listing. All other pharmacies were selected from the yellow pages of a Toledo, Ohio-area
phone book, and were located in northwest Ohio and
southeast Michigan. Each pharmacy was randomly assigned a coded number, and all data was collected
with reference only to those coded numbers.
Data Form
Using OBRA guidelines, the patient counseling information section of Clinical Pharmacology Online,13 and
the patient information section from Drug Facts and
Comparisons,14 we created a checklist-formatted data
form (Fig. 1). We used this form to capture responses
in a dichotomous fashion. That is, for each item on the
checklist, we scored pharmacies as providing the information or service, or not providing it. We weighted
each item equally.
Prescriptions
Five different area physicians wrote a total of 90 prescriptions for hydrochlorothiazide. Each prescription
was for 30 tablets, 25 mg, to be taken once daily with
no refills. Nine separate, written prescriptions were
given to each of 10 monitors. We selected hydrochlorothiazide as the sole medication because (1) it is used
to treat a prevalent condition (eg, hypertension), (2) it
can be used as monotherapy in first-step therapy for
hypertension, (3) it is devoid of abuse potential, (4) it
is used to treat a symptomless condition, and (5) it
is inexpensive.
Filled prescriptions were returned to the University
of Toledo College of Pharmacy for identification, counting, temporary storage, and subsequent destruction.
Monitors
We used a total of 10 monitors to present prescriptions
at the 90 pharmacies, and to complete the data forms
immediately thereafter. Monitors were drawn from a
homogeneous demographic pool of college students.
This step was necessary to assure that nothing about
the presentation of any monitor at any pharmacy
would signal special needs vis-á-vis drug information.
Thus, all monitors were between the ages of 18 to 45
years, and appeared healthy. Moreover, each monitor
was instructed to present himself or herself as taking
no other medications, having no known allergies, and
having no other medical conditions apart from hypertension. Each monitor was assigned to nine different
pharmacies, three in each category (ie, chain, community independent, and on-line). Each monitor was
OBRA 1990
95
Fig. 1. Data form for determining
counseling index.
provided with a total of nine different prescriptions,
and charged with taking each of those to a different
assigned pharmacy. Assignments were not random
because it was intended that monitors were not to
present a prescription at any pharmacy where they
had been a customer or were otherwise known.
Experimental Design
In order to assure inter-rater reliability among the 10
monitors, all monitors received in-service training
about the data form. After this instruction, all monitors were presented with three different vignettes, and
asked to complete the data form for each vignette. One
scenario depicted a pharmacist providing thorough
prescription counseling information, one depicted a
pharmacist providing virtually no counseling information, and one depicted a pharmacist providing an
intermediate amount of counseling information. Individual monitors’ assessments of the quantity and
quality of prescription information in each of these
vignettes was linked to the assessments of the quality
and quantity of prescription information in the field,
thus providing a statistical assessment of the pharmacies corrected for any possible differences in monitors
that may have existed.15
We used a many-facet Rasch model to objectively
assess the ability of each pharmacist to provide
American Journal of Therapeutics (2003) 10(2)
96
SCHATZ ET AL
prescription information based upon the checklist.
This model allows simultaneous quantification of the
ability of each pharmacist (representing each pharmacy), the ease or difficulty of each item within the
checklist, and the rating tendency of each monitor.
The Rasch model used is shown below (Equation 1)
log [PijkⲐ(1 − Pijk)] = Bi − Dj − Mk
(1)
where Pijk is the probability of the ith pharmacist satisfying the jth checklist item as recorded by the kth
monitor, Bi is the ability of the ith pharmacist, Dj is the
ease or difficulty of the jth item on the checklist, and
Mk is the ability of the kth monitor. To estimate these
parameters, it is necessary for all elements of all parameters to be linked.15 To accomplish this task, and to
simultaneously train the monitors on the data form, all
10 monitors were presented with three different vignettes (ideal, adequate, and unsatisfactory prescrip-
tion counseling), and asked to complete the data form
for each vignette. Thus, each monitor became linked to
every pharmacy, even though each monitor completed checklists on nine different pharmacies at most.
A partial matrix created for this linkage is shown in
Figure 2.
Data Analysis
Data was initially fitted to a three-facet Rasch model.15
In this analysis, each item checked on the form represents a pass; each item not checked represents a fail.
The three-facet rating scale model taking the form previously shown (Equation 1) elicits continuous scales
that simultaneously measure item difficulty, overall
pharmacy counseling, and monitor ability. Within the
framework of this model, all items are ranked for difficulty based upon their aggregate pass scores for all
pharmacies, and pharmacies are ranked based upon
Fig. 2. Matrix design for Rasch
analysis. An X indicates that the
monitors recorded something for all
items (1–21) for the test pharmacies
(ie, three vignettes) and the nine assigned (field) pharmacies. The linkage of the monitors, items, and
pharmacists can easily be discerned
in the data matrix.
American Journal of Therapeutics (2003) 10(2)
OBRA 1990
the number of items performed and the post hoc determination of item difficulty built into the Rasch
analysis. Although a Rasch analysis basically omits
items that all pharmacies passed or all pharmacies
failed, it does assign them extreme scores (ie, > 3.0 or
< −3.0). The resultant score for each pharmacy was
denoted as a counseling index score. Model fitting
and related computation of counseling index scores
was accomplished with Facets software version 3.22
(Winsteps, Chicago, IL, USA).16 The analysis provides
a measure of instrument (data form) reliability and
validity.16 The range of counseling index scores could
be expected to fall between +3.0 and −3.0, depending
upon sample size and event rarity for each item. Once
counseling index scores had been determined, the
pharmacy code assignments were broken to enable
the data to be stratified by practice setting (ie, chain,
independent, or on-line). Differences in the counseling
indices between these groupings were analyzed by
an analysis of variance using JMP software (Cary,
NC, USA).17 A Tukey–Kramer HSD post hoc test
was applied if significant differences were found.
Monitor reliability was calculated as the correlation
coefficient for monitors’ responses (ie, assignment of
a 1 if a checklist item had been performed or assignment of a 0 if not performed by a test pharmacy) versus true score of each test pharmacy on each item for
each vignette.
Validity or accuracy of monitors’ ratings was defined by the inter-monitor agreement for each checklist item performed or not performed by each of the
test pharmacies (ie, in each vignette). Specifically, the
95% prediction interval for the slope of this regression
line should embrace unity.18
RESULTS
Seventy pharmacies (77.8%) actually filled prescriptions, and thus could be considered as respondents.
One independent community pharmacy, even though
listed under “pharmacies” in the yellow pages of the
local telephone directory, could not fill the prescription because it sold over-the-counter (OTC) items exclusively. The 19 other nonrespondents were on-line
pharmacies. Thus, prescriptions were filled at all 30
chain pharmacies, 29 community independent pharmacies, and 11 on-line pharmacies. Each of these filled
prescriptions opened the door for a pharmacist–
“patient” counseling experience that could be scored
via the data form. Counseling was provided by on-line
pharmacies either on-line, by e-mail, or by telephone.
On-line pharmacies were initially provided a 1-month
time frame in which to acknowledge receipt of the
97
prescription. If after 1 month there was no response,
monitors were instructed to resubmit the prescription.
If after a second month there was no acknowledgment
of the prescription from an on-line pharmacy, it was
contacted directly by the prescriber’s office. A third
month was allowed to elapse before failure to acknowledge the prescription was considered to be
a nonresponse.
Counseling index scores computed by Rasch analysis are listed in Table 2 by rank.
The more positive scores indicate a greater number
of items were addressed. The more negative scores
indicate fewer items being addressed.
Even without a specific request from monitors,
19 pharmacies (27%) spontaneously initiated counseling. Another 29 of the respondent pharmacies
(41%) actually offered to counsel our monitors without being requested to do so. The minimum requirement of OBRA 1990 is, in fact, that prescription counseling be offered on all prescriptions. This requirement was considered to have been met if monitors
were asked if they wished to speak with a pharmacist,
or if they had any questions, or if, at on-line pharmacies, e-mail or written instructions were provided
about contacting a pharmacist. Thus, prescription
counseling was provided without request by 48 (69%)
of 70 pharmacies. Written pamphlets about hydrochlorothiazide were provided by 55 pharmacies (79%).
Pharmacists were not available for counseling, even
upon request of monitors, at 9 pharmacies (13%). All
prescriptions were filled correctly with 25 mg tablets
of hydrochlorothiazide.
No single item on the data form was performed by
all pharmacies or provided by all pharmacies. Only
two items of information on the data form were not
provided by a single pharmacy. No pharmacy provided drug storage information (item 19), and no
pharmacy provided advice on steps to take in the
event of a missed dose (item 21). The degree of difficulty of the 21 items on the data form ranged from 7.98
(i.e. item 21) to −4.08 (item 3), with larger positive
values representing a higher degree of difficulty, and
larger negative values representing items having been
asked or provided by a large number of pharmacies.
The Rasch estimates of item difficulty ± standard error
(SE), corrected for monitor differences are listed in
Table 1.
Counseling indices ranged from +1.84 to −3.64 with
a reliability of 0.78 (maximum possible 1.0). The mean
indices ± standard error (SE) for each practice setting
were: chain pharmacies, −1.34 (0.18); community independent pharmacies, −1.05 (0.29); and on-line pharmacies, −1.51 (0.36). These means were not significantly different by analysis of variance (ANOVA). The
American Journal of Therapeutics (2003) 10(2)
98
SCHATZ ET AL
Table 1. Rasch item analysis.
Objective
Rasch estimate of
difficulty
SE
Spontaneously provided informaton
Offered prescription counseling
Pharmacist was available for counseling
Written information provided
Checked about other meds
Checked on medical conditions
Checked on allergies to meds
Identified prescription as hydrochlorothiazide
Identified strength as 25 mg
Instructed to take once a day
Instructed to take in AM
Checked work shift
Identified the indications as blood pressure or edema
Warned of increased urination
Warned of photosensitivity
Warned of muscle cramps
Advised potassium supplementation
Advised to monitor blood pressure
Advised on storage conditions
Stipulated no refills
Instructed what to do in case of missed dose
0.37
−0.94
−4.08
−1.88
0.29
0.81
−2.39
−0.88
1.97
−1.76
−1.76
2.58
−0.94
−1.23
.00
1.80
0.54
2.35
7.98
1.80
3.36
0.28
0.24
0.38
0.25
0.28
0.31
0.26
0.24
0.42
0.24
0.24
0.48
0.24
0.24
0.26
0.40
0.29
0.46
1.85
0.40
0.54
Data form item #*
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
*Item numbers correspond with those shown in Figure 1.
†
The more positive the score, the more difficult the item (ie, the less likely that this item would be performed by pharmacies); the more
negative the score, less difficult the item (refer to Equation 1).
SE indicates standard error.
individual variates and their medians are depicted
graphically in Figure 2.
The 2 test for monitor equivalency embedded in
the many-facet Rasch analysis yielded a P value less
than 0.01, suggesting that all monitors were not
equivalent.18 These differences are taken into account,
however, by the many-facet Rasch model. Furthermore, the intermonitor estimate of reliability was 0.88.
The slope value for the estimate of validity (accuracy)
was 0.94 with 95% prediction limits of 0.60 to 1.29.
This suggests a slight bias on the part of monitors to
fail to record an item that was performed, but an adjustment for this bias is automatically included in the
scoring of pharmacies. A 2 test for the variability of
the pharmacies yielded a P value of 0.52, suggesting
that each pharmacy sampled was drawn from a normally distributed population of pharmacies. Figure 3
shows the effect of the practice setting on counseling
index scores.
DISCUSSION
The OBRA 1990 mandate to offer prescription drug
counseling by pharmacists to their Medicaid clients is
American Journal of Therapeutics (2003) 10(2)
Fig. 3. Effect of practice setting on counseling index
scores. Within each Box Plot for chain, independent, and
on-line pharmacies, the individual plotting symbols denote individual pharmacy scores. Lines within box plots
denote medians, and upper and lower box boundaries
depict 75th and 25th percentiles, respectively. For test or
vignette pharmacies, the plotting symbols represent individual pharmacy scores, and the upper and lower box
boundaries depict the maximum and minimum values,
respectively. The solid line across all four different
groups depicts the grand mean score.
99
OBRA 1990
Table 2. Counseling index scores.
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Setting
Index
SE
Rank
Setting
Index
SE
Tank
Setting
Index
SE
Independent
Independent
Independent
Chain
Independent
Independent
Independent
Independent
Independent
Chain
On-line
Independent
Independent
Chain
Independent
On-line
On-line
Chain
Independent
On-line
Independent
Chain
Chain
Chain
1.84
1.66
1.25
0.89
0.54
0.54
0.54
1.19
0.19
0.19
0.02
−0.07
−0.31
−0.31
−0.32
−0.32
−0.32
−0.43
−0.47
−0.48
−0.48
−0.48
−0.48
−0.48
0.64
0.64
0.60
0.59
0.59
0.59
0.59
0.59
0.59
0.59
0.59
0.59
0.58
0.58
0.58
0.58
0.58
0.61
0.58
0.58
0.58
0.58
0.58
0.58
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Chain
Chain
On-line
Independent
Chain
Chain
Chain
Independent
Chain
Independent
On-line
Independent
Chain
Chain
Chain
Chain
Chain
Chain
Independent
Chain
Independent
Chain
Chain
Chain
−0.65
−0.66
−0.82
−0.82
−0.82
−0.88
−1.17
−1.17
−1.18
−1.24
−1.34
−1.34
−1.34
−1.34
−1.34
−1.34
−1.34
−1.34
−1.53
−1.53
−1.56
−1.70
−1.74
−1.91
0.58
0.58
0.58
0.58
0.58
0.59
0.59
0.59
0.61
0.67
0.59
0.59
0.59
0.59
0.59
0.59
0.59
0.59
0.61
0.61
0.63
0.61
0.63
0.63
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
Chain
Chain
Independent
On-line
Independent
Independent
Independent
On-line
Chain
Chain
Independent
On-line
On-line
Independent
Chain
On-line
Independent
Independent
Chain
Chain
Independent
Independent
−1.91
−1.98
−2.08
−2.08
−2.08
−2.08
−2.08
−2.33
−2.48
−2.51
−2.51
−2.83
−3.00
−3.12
−3.12
−3.12
−3.30
−3.35
−3.35
−3.35
−3.64
−3.64
0.63
0.67
0.63
0.63
0.63
0.63
0.63
0.67
0.74
0.67
0.67
0.74
0.74
0.87
0.87
0.87
0.87
1.15
1.15
1.15
0.87
0.87
SE indicates standard error.
a recognition by government that the provision of sufficient prescription information is likely to lower
ADRs and their related health care costs. Some states,
beginning even before the implementation of OBRA
1990, have extended this mandate to all patients, and
this mandate has also been incorporated into Pharmacy’s professional standards. The mandate of OBRA
1990 was implemented in 1993. The study that we
conducted was intended to provide a snapshot of the
extent and quality of counseling on prescription drugs
currently available to consumers. Although the mandate to provide drug information can be technically
met by providing written information, our evaluation
pertained to face-to-face interactions. A report to the
FDA investigators concluded that of the eight informational elements required for written material to
provide complete drug information for patients, only
three were included 60% of the time or more, and that
there was considerable variability in the quantity and
quality of drug information contained therein.19 We
were also interested to know whether the practice setting (ie, community independent pharmacy, large
chain pharmacy, or on-line pharmacy) might influence
any counseling services that are provided. We selected
a total of 90 pharmacies, 30 from each category, as sites
for evaluating counseling services. Our analysis was
limited to 70 respondent pharmacies. Every chain
store pharmacy initially selected was evaluable. Of the
30 independent community pharmacies, 29 were
evaluable. However, only 11 of the 30 on-line pharmacies responded to our prescription submission, even
allowing three attempts over 3 months; therefore,
these were the only evaluable on-line pharmacies.
The data form that we prepared asked whether any
of 21 different informational procedures about hydrochlorothiazide were undertaken at any given pharmacy. A three-facet rating scale model that constitutes
the Rasch analysis yielded a counseling index for each
pharmacy, and simultaneously ranked each item on
the data form for its degree of difficulty. Any of the 21
items on the data form that were either omitted by
every pharmacy or that were implemented by every
pharmacy do not influence the counseling index obtained. The monitors who presented prescriptions at
each pharmacy were between the ages of 18 and 45,
appeared to be in good health, and presented themselves as receiving this prescription for the first time,
having no known allergies, taking no other medications, and having no other medical conditions. The
monitors received training on the completion of the
American Journal of Therapeutics (2003) 10(2)
100
data forms, and when tested on their scoring of three
different vignettes, their rankings were highly correlated. Differences between monitors were modeled in
the many-facet Rasch model, which, in fact, may be an
advantage to this approach. Monitors were instructed
to complete data forms immediately after receiving
their prescriptions (or their prescription information
from on-line pharmacies) to reduce recall bias.
Practice setting did not influence the average counseling index. These values averaged −1.33, −1.51 ,and
−1.05 for large chain, on-line, and community independent pharmacies, respectively. These differences
were not statistically significant, and were consistent
in that all practice settings exhibited, on average, comparably adequate counseling indices. Provision of
relatively thorough drug information is not out of the
reach of community pharmacies. Eleven pharmacies
exhibited high scores (>0), and eight of those (73%)
were community independent pharmacies, two (18%)
were large chain pharmacies, and one (9%) was an
on-line pharmacy. Only 27% of the pharmacies spontaneously counseled our monitors, and this is in line
with the frequency of spontaneous counseling reported by others.20
This study has several limitations. We studied only
a limited number of pharmacies in northwest Ohio
and southeast Michigan. It is impossible to know
whether these findings would be reproduced elsewhere in the US. Additionally, we evaluated the prescription counseling for a single drug, hydrochlorothiazide, under the least complicated conditions of use. It
cannot be known whether the likelihood of pharmacist engagement in more extensive prescription counseling activity would have been greater had the patients’ condition or the drug regimen been more complicated. There is no consensus about those elements
of drug information that might be viewed as essential
or “ideal.” Just how much information might be necessary to reduce ADEs is an open question. On the
other hand, the inclusion on our data form of drug
information about hydrochlorothiazide that may be
regarded by some as either unnecessary or as too
much detail for customary counseling neither prejudges nor affects the ability of any pharmacy to
achieve a high counseling index because Rasch analysis simply omits from its computation of individual
counseling indices any of those counseling items that
are simply not undertaken by any pharmacy.
Among the pharmacies in this study, 69% actually
provided prescription counseling. This may signal an
improvement over the extent of counseling provided
just 3 to 4 years ago. In the greater Kansas City Missouri area, only 30% of pharmacies provided prescripAmerican Journal of Therapeutics (2003) 10(2)
SCHATZ ET AL
tion counseling.21 The experience with community
pharmacies in Kansas City and also in Nebraska12 (described previously) may have been fairly representative of all pharmacies, including those in acute care
settings. A national clinical pharmacy services study
conducted in 1998 reported that between 40% and 50%
of hospital pharmacies provided prescription counseling to inpatients either during hospitalization or at
discharge,22 whereas another analysis undertaken at
hospitals in Columbus, Ohio indicated that fewer than
one-third of hospital pharmacists counseled patients.23 Though at first blanche these rates of prescription counseling may appear low, they do signal a clear
improvement over those observed before the implementation of OBRA 1990. For example, a study of 80
community pharmacists in New Mexico that was undertaken just before the implementation of OBRA
1990 found that prescription counseling was provided
with about 20% of the prescriptions. For patients
whose prescriptions were paid for by capitation, the
counseling rate dropped to about 8%.24 Thus, the extent of counseling offered seems to be trending upward as follows: 10% to 20% pre-OBRA, 30% to 50% in
the late 1990s, and ∼70% currently.
If our assessment of prescription counseling for hydrochlorothiazide can be extrapolated, it would suggest that the objectives of OBRA 1990 on the quality of
counseling services provided by pharmacies are adequately being met.
Ultimately, however, the issue of prescription counseling effectiveness must be addressed by outcome
studies. This conclusion is supported by the findings
of others. For example, an analysis by the FDA of 12year trends in prescription counseling both in pharmacies and in physician offices has likewise led to the
conclusion that there has been moderate progress on
verbal counseling stemming from OBRA 1990.25 On
the other hand, a survey of community pharmacy clients with asthma indicated that community pharmacists rarely provided information regarding either the
role of their asthma medications, techniques for appropriate inhaler use, or general advice regarding
asthma control.26 Taken together, these findings suggest that prescription counseling can generally be
characterized as adequate or moderate, though not
good or excellent. This should be cause for concern,
particularly when many studies that have evaluated
outcomes in patient safety and well-being with direct
pharmacist involvement in the therapeutic regimen,
have demonstrated benefits.27–31 In-house counseling
of hospitalized patients receiving combination lipidlowering therapy after coronary artery bypass graft
101
OBRA 1990
surgery coupled with weekly pharmacist-patient telephone contact for the next 12 weeks dramatically increased patient compliance measured both 1 and 2
years later.32
The quality of counseling services rendered by
pharmacists in Great Britain has also been criticized,
though in these studies it was advice pertaining to
over-the-counter (OTC) drugs that was evaluated.33
Canadian pharmacists do not seem particularly eager
to become more engaged in providing information
about OTC drugs, either.34
Whether the barriers that limit pharmacist advicegiving on OTC drugs in Great Britain and Canada are
similar to those that might be limiting prescription
counseling in US pharmacies remains to be determined; however, a substantial number of potential
barriers to widespread prescription counseling in the
U.S. have begun to be identified. For example, a survey of asthma patients revealed that only 36% even
desired any information on potential druginteractions or side effects, whereas 33% believed that
drug information should be provided by their physicians.25 When these same patients were asked whether
they would be willing to pay for prescription counseling or to have their insurance pay for it, 76% indicated
that they would not be willing to pay and 46% indicated that they would not be willing to have their
insurance pay for this service. In another survey of
pharmacy provision of asthma patient care, pharmacists were viewed by patients as not substantially contributing to improved control of their condition.35 If
willingness to pay is used as a gauge of consumer
desire for pharmacy services—including prescription
counseling—then a few recent surveys suggest a lack
of demand for these services from pharmacists. Metge
et al reported that most patients would accept pharmacy services only if they were free.36 Others reported
similar findings.37 In contrast, 96.7% of surveyed consumers (n = 197) indicated a willingness to pay for
counseling services, but valued this service at $2 to
$3.38 More recently, an overall (out-of-pocket) willingness-to-pay figure for pharmacy services was established to range from about $4.00 to $5.50 per prescription.39 Though drug information from pharmacists
may not be a highly valued commodity, most adult
outpatients do want to learn from their physicians
about adverse drug reactions.40
Existing workload exigencies may be responsible
for an inertia that is difficult to overcome in common
practice settings.41 Prescription counseling accounted
for 6.6% of pharmacists’ work-related functions in a
grocery chain, or about 30 minutes per shift, with nonjudgmental functions dominating the workday. 42
Thus, a perception by pharmacists that prescription
counseling is not highly valued by consumers,43
coupled with lack of compensation for those services,
the difficulty of overcoming the inertia of extant work
routines including a lack of time, staffing, and space to
accommodate this function, might all serve as barriers
to a more active role for pharmacists in prescription
counseling.44,45 Even though much of the information
about drugs placed on consumer-targeted websites is
accurate, consumers have difficulty in interpreting the
information, and reliable sources of information are
still required.41 Safety-minded statutory changes such
as those in OBRA 1990 might be more likely to effectively meet their objectives if they include provisions
for reimbursement for fundamentally new services.
CONCLUSIONS
In conclusion, we found that when presented with a
nonrefillable prescription for oral hydrochlorothiazide
(25 mg once a day), pharmacies generally provided
adequate prescription counseling services. This was
true whether the pharmacies were community independent, large chain, or on-line pharmacies. Sixty-nine
percent of these pharmacies met minimum state standards for counseling by offering to provide prescription counseling. Nine pharmacies (13%) declined to
make a pharmacist available on request, and this probably was a function of manpower shortages (too few
pharmacists per client) rather than an unwillingness to
provide prescription counseling. The Rasch analysis
that we used yielded counseling index scores that simultaneously provided both a quantitative and qualitative assessment of the prescription counseling function. This analysis did not require any subjective ranking of the service whatsoever, and the outcome is
therefore free from rater or investigator biases. The
Nominal Group Technique is one of several more subjective methods developed for assessing other types of
health care delivery that is currently being used in the
evaluation of pharmacy counseling services, too.46
This study was not designed to assess outcomes associated with prescription counseling. Because of the
limitations of our study, our findings can only be considered as highly preliminary. But if they are validated
by similar protocols or by other strategies, we will
have to address the structural problems in the delivery
of prescription counseling and reimbursement thereof
in a much more concerted way if we are likely to have
any appreciable success in reducing the adverse consequences tied to prescription drug use, as was envisioned when OBRA 1990 was fashioned.
American Journal of Therapeutics (2003) 10(2)
102
Although the objectives of OBRA 1990 vis-á-vis prescription counseling services appear to be adequately
met, only proper outcomes research can establish
whether the objectives of reduced ADEs and related
health care costs are also being met.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the assistance
of Drs. Monica Holiday-Goodman, James Tita, Luis
Jauregui, Vijay Mahajan, Albert Rafanan, and Srinivas Katragadda.
REFERENCES
1. Johnson JA, Bootman JL: Drug-related morbidity and
mortality. A cost of illness model. Arch Intern Med 1995;
155:1949–1956.
2. Hepler CD, Strand LM: Opportunities and responsibilities in pharmaceutical care. Am J Hosp Pharm 1990;47:
533–543.
3. Col N, Fanale JE, Kronholm P: The role of medication
noncompliance and adverse drug reactions in hospitalizations of the elderly. Arch Intern Med 1990;150:841–
845.
4. Dennehy CE, Kishi DT, Louie C: Drug-related illness in
emergency department patients. Am J Health-System
Pharm 1996;53:1422–1426.
5. Lakshmanan MC, Hershey CO, Breslau D: Hospital admissions caused by iatrogenic disease. Arch Intern Med
1986;146:1931–1934.
6. Bates DW, Cullen DJ, Laird N, et al: Incidence of adverse
drug events and potential adverse events. Implications
for prevention. ADE Prevention Study Group. JAMA
1995;274:29–34.
7. Willcox SM, Himmelstein DU, Woolhandler S: Inappropriate drug prescribing for the community-dwelling elderly. JAMA 1994;272:292–296.
8. Aparasu RR, Sitzman SJ: Inappropriate prescribing for
elderly outpatients. Am J Health Syst Pharm 1999;56:
433–439.
9. Stuck AE, Beers MH, Steiner A, et al: Inappropriate
medication use in community-residing older persons.
Arch Intern Med 1994;154:2195–2200.
10. Pub L No. 101–508, §4401, 1927(g) (November 5, 1990);
OBRA 1990 Regulations, Federal Register, November 2,
1992;57FR(212):4937–4940.
11. Molzon JA: What kinds of patient counseling are required? Am Pharm 1992;NS32:50–57.
12. Scott DM, Wessels MJ: Impact of OBRA ’90 on pharmacists’ patient counseling practices. J Am Pharm Assoc
(Wash) 1997;NS37:401–406.
13. Clinical Pharmacology 2000. Patient Education. Available at: http://www.cp.gsm.com. Accessed
American Journal of Therapeutics (2003) 10(2)
SCHATZ ET AL
14. Drug Facts and Comparisons, Wolters Kluwer Co., St.
Louis, 2000, pp 619.
15. Linacre JM: Many-Facet Rasch Measurement, 2nd edition.
MESA Press, Chicago, 1994.
16. Linacre JM: A User’s Guide to Facets, Version 3.0. MESA
Press, Chicago, 1996.
17. JMP Statistics and Graphics Guide, version 3.1 of JMP. SAS
Institute, Inc. Cary, NC, 1995.
18. Montgomery DC, Peck EA: Introduction to linear regression analysis. John Wiley and Sons, Inc., New York, 1982.
19. Svarstad B, Mount JK: Evalutation of written prescription information provided in community pharmacies,
2001 [Center for Drug Evaluation and Research Web
site]. December 21, 2001. Available at: http://www.fda.
gov/cder/reports/prescriptionInfo/default.htm. Accessed June 19, 2002.
20. Willison DJ, Muzzin LJ: Workload, data gathering, and
quality of community pharmacists’ advice. Med Care
1995;33:29–40.
21. Fritsch MA, Lamp KC: Low pharmacist counseling rates
in the Kansas City, Missouri, metropolitan area. Ann
Pharmacother 1997;31:984–991.
22. Raehl CL, Bond CA: 1998 national clinical pharmacy services study. Pharmacotherapy 2000;20:436–460.
23. Griffith NL, Schommer JC, Wirsching RG: Survey of inpatient counseling by hospital pharmacists. Am J Health
Syst Pharm 1998;55:1127–1133.
24. Raisch DW: Patient counseling in community pharmacy
and its relationship with prescription payment methods
and practice settings. Ann Pharmacother 1993;27:1173–
1179.
25. Morris LA, Tabak ER, Gondek K: Counseling patients
about prescribed medication: 12-year trends. Med Care
1997;35:996–1007.
26. Liu MY, Jennings JP, Samuelson WM, et al: Asthma patients’ satisfaction with the frequency and content of
pharmacist counseling. J Am Pharm Assoc (Wash) 1999;
39:493–498.
27. Brouker ME, Gallagher K, Larrat EP, et al: Patient compliance and blood pressure control on a nuclear powered aircraft carrier: impact of a pharmacy officer. Mil
Med 2000;165:106–110.
28. Berringer R, Shibley MC, Cary CC, et al: Outcomes of a
community pharmacy-based diabetes monitoring program. J Am Pharm Assoc (Wash) 1999;39:791–797.
29. Currie JD, Chrischilles EA, Kuehl AK, et al: Effect of a
training program on community pharmacists’ detection
of and intervention in drug-related problems. J Am
Pharm Assoc (Wash) 1997;NS37:182–191.
30. Faulkner MA, Wadibia EC, Lucas BD, et al: Impact of
pharmacy counseling on compliance and effectiveness
of combination lipid-lowering therapy in patients undergoing coronary artery revascularization: a randomized, controlled trial. Pharmacotherapy 2000;20:410–416.
31. Birtcher KK, Bowden C, Ballantyne CM, et al: Strategies
for implementing lipid-lowering therapy: pharmacybased approach. Am J Cardiol 2000;85:30A-35A.
32. Faulkner MA, Wadibia EC, Lucas BD, et al: Impact of
OBRA 1990
33.
34.
35.
36.
37.
38.
pharmacy counseling on compliance and effectiveness
of combination lipid-lowering therapy in patients undergoing coronary artery revascularization: a randomized, controlled trial. Pharmacother 2000;20:410–416.
Bissell P, Ward PR, Noyce PR: Appropriateness measurement: application to advice-giving in community
pharmacies. Soc Sci Med 2000;51:343–359.
Taylor J, Berger B, Anderson-Harper H, et al: Pharmacists’ readiness to assess consumers’ over-the-counter
product selections. J Am Pharm Assoc (Wash) 2000;40:
487–494.
Kradjan WA, Schulz R, Christensen DB, et al: Patients’
perceived benefit from and satisfaction with asthmarelated pharmacy services. J Am Pharm Assoc (Wash)
1999;39:658–666.
Metge CJ, Hendricksen C, Maine L: Consumer attitudes,
behaviors, and perceptions about pharmacies, pharmacists, and pharmaceutical care. J Am Pharm Assoc
(Wash) 1998;38:37–47.
Smith DL: APhA national survey: willingness of consumers to pay for pharmacists’ clinical services. Am
Pharm 1983;23:58–64.
Carroll NV, Perri M 3rd, Eve EE, et al: Estimating demand for health information: pharmacy counseling services. J Health Care Mark 1987;7:33–40.
103
39. Suh DC: Consumers’ willingness to pay for pharmacy
services that reduce risk of medication-related problems.
J Am Pharm Assoc (Wash) 2000;40:818–827.
40. Ziegler DK, Mosier MC, Buenaver M, et al: How much
information about adverse effects of medication do patients want from physicians? Arch Intern Med 2001;161:
706–713.
41. Curran CF, Oh KE: Sources of drug information available to consumers. Drug Information Journal 2001;35:
539–546.
42. Dupclay L Jr, Rupp MT, Bennett RW, et al: Analysis of
grocery chain pharmacists’ work-related behaviors. J
Am Pharm Assoc (Wash) 1999;39:74–81.
43. Schommer JC, Wiederholt JB: Pharmacists’ perceptions
of patients’ needs for counseling. Am J Hosp Pharm
1994;51:478–485
44. O’Loughlin J, Masson P, Dery V, et al: The role of community pharmacists in health education and disease prevention: a survey of their interests and needs in relation
to cardiovascular disease. Prev Med 1999;28:324–331.
45. Raisch DW: Barriers to providing cognitive services. Am
Pharm 1993;NS33:54–58.
46. Ward PR, Bissell P, Noyce PR: Criteria for assessing the
appropriateness of patient counseling in community
pharmacies. Ann Pharmacother 2000;34:170–175.
American Journal of Therapeutics (2003) 10(2)