- Email: [email protected]
Symptomatology, attribution to medicines, and symptom reporting among Medicare enrollees
Available online at www.sciencedirect.com
Research in Social and Administrative Pharmacy 5 (2009) 225–233
Original Research
Symptomatology, attribution to medicines, and symptom reporting among Medicare enrollees Olayinka Oladimeji, B.Pharm., Karen B. Farris, B.S.Pharm., Ph.D.*, Julie G. Urmie, Ph.D., William R. Doucette, Ph.D. Department of Pharmaceutical Socioeconomics, University of Iowa, Iowa City, IA 52242, USA
Abstract Background: Adverse drug events (ADEs) are generally preceded by symptoms reported by patients, and understanding who reports symptoms is important in understanding ADEs. Objectives: The objectives of this study were to (1) determine the prevalence of symptomatology experienced, recalled, and reported to physicians (2) quantify the extent of attribution of recalled symptoms to medicines, and (3) predict factors associated with the reporting of recalled symptoms. Methods: This was an Internet-based survey administered by Harris InteractiveÒ. Data collected and included in the analysis were symptoms experienced, symptom attribution to medicines, number of medicines, concern and necessity beliefs in medicines, self-rated health, number of physician visits, hospitalizations and emergency room visits in the past 6 months, whether patients had to pay part of their prescription costs, number of pharmacies, self-reported medication adherence, and sociodemographics. A descriptive analysis of sociodemographic and clinical/behavioral characteristics was completed, the frequency distribution of symptoms and their attribution to medication was obtained, and independent factors that predicted who was likely to report symptoms to physicians, using a multiple logistic regression analysis, were identified. Results: Sixty-two percent (n ¼ 751) reported having at least 1 of 10 symptoms in the past month. Of the 1220 Medicare enrollees, 6.0-31.1% experienced varied specific symptoms. Of those who recalled they had experienced a particular symptom, 11.7-48.8% thought the symptoms were related to the medicine they were taking. Reporting symptoms to physicians was positively related to concern beliefs, seeing physicians R 2 times in the past 6 months, perceiving symptom experienced was due to medicines, using more than 3 pharmacies, and not having to pay part of prescription costs. Conclusions: Assessing beliefs in medications may help to identify unreported symptoms and subsequent ADEs. In addition, symptom reporting to physicians and other health professionals should be encouraged so that preventable ADEs are detected. Ó 2009 Elsevier Inc. All rights reserved. Keywords: Symptom reporting; Adverse drug events; Medicare
* Corresponding author. Division of Clinical and Administrative Pharmacy, S-525 Pharmacy Building, 115 S. Grand Avenue, Iowa City, IA 52242-1112, USA. Tel.: þ1 319 384 4516; fax: þ1 319 353 5646. E-mail address: [email protected] (K.B. Farris). 1551-7411/09/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.sapharm.2008.08.004
226
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233
Introduction Adverse drug events (ADEs) are generally preceded by symptoms reported by patients, and the process by which symptoms are labeled may be applicable to the attribution of ADEs.1 Symptoms serve as triggers for cognitive processes, and the ability of patients to identify a symptom as possibly drug-induced is tied closely to the process by which patients label them.2,3 Symptoms are, therefore, the basis for identifying an ADE but not the definition of it, which is an injury resulting from medical interventions related to the use of a drug.4,5 Patients may have symptoms for months without any changes in their medications, and only a small percentage of patients report that symptoms actually led to a physician visit.6 However, Weingart et al reported that patients discussed 196 of their 286 medication symptoms with their physicians and 22% of the symptoms were serious enough to require a visit to a medical facility. Twenty-three percent of patients’ unreported medication symptoms led to preventable or ameliorable ADEs. Overall, 65% of the cases had ameliorable ADEs, which occurred as a result of failure of physicians to act on self-reported symptoms. They also reported that for every symptom that patients experienced but failed to report, 1 in 5 resulted in an ADE that could have been prevented or been made less severe. They did not inquire into patient’s reasons for not reporting some of the adverse side effects they experienced. It was speculated that patients may have been reticent about discussing side effects that were either common or relatively mild or that were embarrassing such as diarrhea or sexual dysfunction. Thus, patient-reported medication symptoms represent a valuable source of information about medication safety.7 Various factors have been identified in the reporting of symptoms. Patients who take more medications and have more drug allergies are more likely to report medication symptoms to their physicians. These patients may also have more physician encounters and, therefore, more opportunities to report symptoms.7 Patients with fewer symptoms are probably able to manage and tolerate them, whereas those with more symptoms would rather seek the help of a health professional. Patients who report more symptoms are also more likely to report an ADE to their physicians.8 The common sense model of illness is a framework that attempts to explain how people may adapt to and/or manage health threats such as
symptoms. This model may be useful to examine the representation of symptoms experienced by Medicare enrollees and their coping strategy for dealing with it, reporting symptoms to physicians. The common sense model of illness views individuals as active problem solvers who strive to assign meaning to a somatic sensation, such as symptoms, and this typically involves a representation of the symptom and a coping procedure for dealing with the symptom.9 There are typically 5 distinct attributes of illness representations that may be evoked when a person experiences a symptom, including identity, timeline, cause, control, and consequences. With reference to reporting of symptoms and their attribution to medications and to ADEs, individuals may consider all of these attributes.9-11 In terms of measuring attribution to medications, perceived attribution of a symptom to a medication has been assessed using a constructed scale, the medication attribution scale (MAS). Aversa et al found that negative attributions about the effects of anti-retroviral medications were related to increased depressive symptoms, discontinuation of therapy, and quality of life.11 Weingart et al measured perceived medication attribution with a single item.7 Horne et al used the common sense model of illness to show that necessity and concern beliefs about medicines are 2 core themes that people use when interpreting symptoms and causal attributions related to their medications.12,13 Necessity beliefs relate to an individual’s perception of the necessity of medication for maintaining health, whereas concern beliefs relate to an individual’s concern about the adverse consequences of medication based on beliefs about the potential for dependence or harmful long-term effect. Items from the necessity scale include, for example, ‘‘My health at present depends on my medicines’’ and ‘‘Without medicines, I would be very ill.’’ Concern beliefs comprise items such as ‘‘I sometimes worry about the long-term effects of my medicines,’’ ‘‘Having to take my medicine worries me,’’ and ‘‘My medicines disrupt my life.’’ Significant relationships between medication beliefs and adherence have been reported, with higher scores on the concern-belief scale correlating with lower reported adherence. In addition, higher reported adherence rates were associated with higher necessity-concerns difference scores. The concern-belief scale, therefore, assesses the negative attitude of individual patients toward medication, whereas the necessity-belief scale assesses their positive attitude toward medication.12,13
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233
It is hypothesized that patients with higher scores on the concern-beliefs scale may be more cognitively aware of their symptoms and pay particular attention to unwanted reactions and are, therefore, likely to report symptoms. Higher scores on the scale indicate stronger negative attitudes toward taking medication, not positive beliefs, because the individuals worry more about the long-term effects of their medicines and their dependency on their medicines. In addition, it is hypothesized that patients with higher scores on the necessity-beliefs scale would be more likely to report symptoms because they are more likely to take their medications and may monitor the effects, making them more aware of reactions. At the same time, the opposite could occur for necessity beliefs. For example, patients could be more tolerant of side effects because they value the necessity of continuing their therapy and may be less likely to report their symptoms. The objectives of this study were to (1) determine the prevalence of symptomatology experienced, recalled, and reported to physicians (2) quantify the extent of attribution of recalled symptoms to medicines, and (3) predict factors associated with the reporting of recalled symptoms.
Methods Design An Internet-based survey was administered by Harris InteractiveÒ on behalf of the University of Iowa, College of Pharmacy. University of Iowa investigators designed the survey and analyzed the data. The project was approved by the Institutional Review Board of the University of Iowa. Patients/setting Harris InteractiveÒ maintains a confidential panel of individuals who have agreed to be invited to participate in telephone and/or online surveys. Harris InteractiveÒ invited individuals in their online panel to participate in this study, and participants received credit from Harris InteractiveÒ for completing the online survey. The inclusion criteria were being 65 years or older, English speakers, US residents, and enrolled in Medicare. Harris InteractiveÒ used their online panel of potential subjects and provided data to University of Iowa researchers from a non-probability sample of 1220 anonymous respondents.
227
Data collection Respondents completed a 166-item survey, which was administered between October 13 and October 19, 2005. The survey took approximately 18 minutes to complete, as numerous skip patterns were included for questions that did not apply to some respondents. In addition, respondents could answer part of the survey and return later if necessary for completion. The survey was pilot tested in Iowa City using a convenience sample of 30 older adults at the local senior citizens center. These individuals completed the survey and provided comments about the clarity of survey questions. Also, students and colleagues completed the online version before its release. Information on health symptoms experienced and recalled in the past month (yes/no) was collected. The ‘‘past month’’ time frame was used to improve recall. The symptoms measured were headaches, dizziness, or problems with balance; stomach or gastrointestinal problems; muscle aches, incontinence, or problems with urinating; rash or itching; problems with sleep; changes in mood, fatigue, and sexual problems. The predetermined list of symptoms was based on a previous study that identified ADEs.7 There was no opportunity for respondents to report other nonlisted symptoms. Respondents were also asked if they thought the symptoms they experienced were related to a medicine they were taking.7 Next, they were asked if they reported the symptom to their physician. If no, they were asked if they intended to report the symptom to their physician in the future. Data used in this analysis included age, race, sex, education, income, number of medicines used on a regular basis, concern and necessity beliefs in medicines, self-rated health, number of physician visits in the past 6 months, number of hospitalizations in the past 6 months, number of emergency room visits in the past 6 months, whether they had to pay part of their prescription costs, number of pharmacies where they got their prescription medicines, and self-reported medication adherence. For sociodemographic data, the age of respondents, racial background, sex, and the highest level of education completed were determined. Respondents were asked about their self-rated health status. Respondents rated their health compared with other individuals of their own age, and a 5-item response scale anchored with poor and excellent was used.14,15 To determine the number of medications used, the number of different prescription medicines
228
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233
used in the past month was asked. Then, respondents indicated the number of medications they took on a regular basis, among those they had taken in the past month. To examine their concerns and necessity beliefs about their medications, the 10 items from Horne et al scale were used.11 Five items ask about concern beliefs and include items such as ‘‘Having to take medicines worries me’’ and ‘‘I sometimes worry about the long-term effects of my medicines.’’ Necessity beliefs were assessed with 5 items, including, for example, ‘‘My life would be impossible without my medicines’’ and ‘‘Without medicines, I would be very ill.’’ Five-point Likert scales anchored with strongly disagree and strongly agree were used. Items for concern and necessity beliefs were summed separately, providing a scale with a range of 5-25 for either scale. Previous studies using these scales show reliability estimates ranging from 0.65 to 0.86 and its construct validity was established using factor analysis and hypothesis testing.11 The respondents stated the number of times they had seen a physician, visited the hospital, or been admitted to the emergency room in the past 6 months. They also indicated whether they had to pay part of their prescription drug costs. Respondents indicated the number of pharmacies where they had obtained their prescription medicines in a typical month. Medication adherence was measured using 4 items from the Morisky scale.16 Questions such as ‘‘During the past month, have you ever forgotten to take your medication?’’ and ‘‘When you feel better, do you sometimes stop taking your medication?’’ were answered by the respondents. These 4 items were summed in the analyses. Analysis All analyses were completed by investigators at the University of Iowa. A descriptive analysis of sociodemographic and clinical/behavioral characteristics was completed (Table 1). The frequency distribution of symptoms, respondents’ attribution to medication, and the percentage of those who reported each symptom to physicians were determined (Table 2). Independent factors that predicted who was more likely to report symptoms to physicians, using a multiple logistic regression analysis to obtain odds ratios (ORs) with 95% confidence intervals, were identified (Table 3). The dependent variable was reporting at least 1 symptom to
physicians, among individuals who experienced any symptom in the past month. Sociodemographic data, clinical characteristics, and behavioral characteristics were used as independent variables in the regression. Most independent variables were recoded into dummy variables. Respondents who had not seen a physician in the past 6 months were excluded because they were unlikely to be able to report symptoms to physicians in the ‘‘past month.’’ Also, 6 months was thought to be an appropriate recall period for utilization of health services. Those who had missing data from other variables were also excluded from the analysis. Statistical analyses were performed using SPSS software (version 15.0). Results Participants were between 65 and 94 years old, and 54.9% were female. Most respondents were white, had some college experience, used more than 1 prescription medicine on a regular basis, obtained their prescription medicines from 1 pharmacy, and perceived themselves to be relatively in good health (Table 1). Sixty-two percent (n ¼ 751) reported having at least 1 of the 10 symptoms in the past month. Of the 1220 Medicare enrollees, a range of 6.0-31.1% experienced a particular symptom (Table 2). Of those who stated they had experienced a particular symptom, 11.7-48.8% thought the symptoms were related to the medicine they were taking. A range of 34.3-55.2% reported the particular symptoms to their physicians, depending on the symptom. Consistent with previous literature, patients reported symptoms related to muscular aches, fatigue, sleep problems, sexual problems, and gastrointestinal problems (43.2%, 48.1%, 44.4%, 48.8%, and 50.4%, respectively) more often to their physicians than headache and incontinence (34.3% and 41.0%).5 Patients were least likely to report headaches. Sixty percent of individuals having at least 1 of the 10 symptoms in the past month reported 1 or more of their symptoms to their physicians. About two thirds of individuals with an unreported symptom had an intention to report at least 1 symptom. In the full model, reporting symptoms to physicians was related to concern beliefs about medications (OR ¼ 1.08, 95% CI ¼ 1.03-1.14), seeing physicians R2 times in the past 6 months (OR ¼ 4.07, 95% CI ¼ 1.5210.93), perceiving symptom experienced was due to a medicine (OR ¼ 2.71, 95% CI ¼ 1.76-4.17),
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233 Table 1 Descriptive characteristics of the study populationa (n¼1220)
Variable
Number (%)
Sociodemographics Age (y), mean 65–74 75–84 R85
746 (61.1) 441 (36.1) 31 (2.5)
Sex Male Female
543 (45.1) 661 (54.9)
Racial background White Hispanic African American Other
Mean Standard deviation 72.89 5.68
1106 59 18 25
(90.7) (4.8) (1.5) (2.3)
Highest level of education Less than high school (HS)/had HS degree Some college Had a college degree Had a graduate degree
420 (34.5) 199 (16.3) 209 (17.1)
Annual household income Less than $15,000 $15,000 to $24,999 $25,000 to $34,999 $35,000 to $49,999 $50,000 to $74,999 O75,000
81 196 262 270 191 149
199 (16.3)
(6.6) (16.1) (21.5) (22.1) (15.7) (12.2)
Clinical characteristics Self-rated health Excellent Very good Good Fair Poor Number of medicines 0 1–3 4–6 7–24 Self-reported adherence 0 1 2
151 423 439 184 20
(12.4) (34.8) (36.1) (15.1) (1.6)
133 456 284 257
(10.9) (37.4) (23.3) (21.1)
Table 1 (Continued )
Variable
Number (%)
592 (48.5) 286 (23.4) 152 (12.5) (Continued)
Mean Standard deviation
3 26 (2.1) 4 5 (0.4) Concern beliefs of 11.96 3.97 medicines (ranges from 5-25), mean SD (lower scores is less concern) Necessity beliefs of 16.62 4.81 medicines (ranges from 5-25, mean SD (higher scores is more beliefs) Times seen a physician in the past 6 mo 0 116 (9.6) 1 302 (24.9) 2 400 (33.0) 3 or more times 394 (32.5) Times visited the emergency department in the past 6 mo 0 1054 (86.8) 1 132 (10.9) 2 23 (1.9) 3 or more times 6 (0.5) Times admitted in the hospital in the past 6 mo 0 1040 (85.2) 1 147 (12.1) 2 24 (2.0) 3 or more times 5 (0.4) Behavioral characteristics Do you have to pay part of Yes, pays part of costs No, pays none of costs Has no prescription insurance Number of pharmacies 0 1 2 3–10
Geographic region (state of residence)b North East 257 (20.9) Midwest 393 (32.2) South 319 (26.1) West 248 (20.4)
229
your prescription costs? 467 (62.2) 42(5.6) 224 (29.8)
127 852 201 39
(10.4) (69.9) (16.5) (3.2)
a Values are number (percentage) except otherwise indicated. Numbers that do not add up to 1220 indicate missing data. b The North East region includes Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, and Vermont. Midwest region includes Iowa, Indiana, Illinois, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, and Wisconsin. South region includes Alabama, Arkansas, Delaware, District of Columbia, Florida, Georgia, Kentucky, Louisiana, Maryland, North Carolina, Oklahoma, South Carolina, Tennessee, Texas, Virginia, and West Virginia. West region includes Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, and Wyoming.
230
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233
Table 2 Self-reported symptoms in the past month among medicare enrollees (n ¼ 1220)a Symptoms
Experienced any symptom n (%)a
Related it to medicine n (%)b
Reported symptom to physician n (%)b
Muscle aches Sleep problems Fatigue Stomach problems Dizziness Incontinence Headache Rash itching Sexual problems Mood changes
380 367 337 234 192 195 172 140 125 73
56 43 75 60 66 23 24 40 60 24
164 163 162 118 106 80 59 71 61 30
a b
(31.1) (30.1) (27.6) (19.2) (15.7) (15.8) (14.1) (11.5) (10.2) (6.0)
(14.7) (11.7) (22.2) (25.6) (34.4) (11.8) (24.0) (28.6) (48.0) (32.9)
(43.2) (44.4) (48.1) (50.4) (55.2) (41.0) (34.3) (50.7) (48.8) (41.1)
Denominator in column 2 is 1220. Denominator in columns 3 and 4 is the frequency of experiencing that specific symptom (number shown in column 2).
using more than 3 pharmacies (OR ¼ 3.60, 95% CI ¼ 1.10-11.13), and not having to pay part of the prescription costs (OR ¼ 3.80, 95% CI ¼ 1.5711.13) (Table 3). Discussion Understanding who reports symptoms may help us understand ADEs and their reporting. In this study, 62% of individuals reported having at least 1 symptom in the past month and a range of 6.0-31.0% experienced a particular symptom. Concern beliefs in medicines, having 2 or more physician visits, using more than 3 pharmacies, and perceiving symptom experienced was due to medicines were associated with reporting symptoms to physicians. Necessity beliefs about medicines were not related to symptom reporting. The attributes of symptoms and the resulting coping strategies to manage symptoms should align.9 For example, if a patient experiencing a symptom can label, attribute its cause to a medicine rather than age or a medical condition, and believe cure/control of the symptom can be achieved by seeking the help of a health provider, the patient may be likely to report the symptom to a health provider. The results showed that over half the respondents experienced at least 1 symptom in the past month; however, the attribution to medications did not seem high. Aversa et al, using the MAS, showed that about 39-52% respondents attributed the symptoms they were experiencing to the medicine they were taking.11 Also, in a previous study, it was found that about 18% of elderly patients reported an unwanted reaction to the medicine they were taking.8 From these results, it can be inferred that some of the
attributions to medications were lower than expected. Lower than expected attribution rates may have occurred because symptoms were not caused by medications. Respondents may also have attributed symptoms to other causes such as disease or age, or respondents may have lacked sufficient information and/or experience to attribute the symptom to medications. Consistent with the findings from Weingart et al., it is important for health providers to verify symptoms that patients may perceive as normal or due to aging to be sure they are not related to the medicine they are taking.7 As hypothesized, stronger concern beliefs about medicines were positively related to reporting symptoms to physicians, although the ORs were modest. People with stronger concern beliefs might be more sensitive to their symptoms and more watchful for unwanted reactions and thus more likely to report symptoms. Contrary to the hypothesis, necessity beliefs in medicines were not significantly related to symptom reporting. Necessity beliefs in medicines are important in adherence but may not be important in symptom reporting because individuals believe that they are dependent on their medicines and, therefore, may be more willing to tolerate symptoms that do arise and thus not report symptoms to their physicians. Respondents were not asked if they had experienced the particular symptoms in the past. If so, respondents may have learned previous coping strategies for managing symptoms. An important issue with these beliefs is that they may not be generalizable to all patients’ medications, and this inconsistency in specificity may have contributed to the low OR for concern beliefs and the nonsignificant finding for necessity beliefs.
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233 Table 3 Logistic regression predicting reporting of symptoms to physicians (n ¼ 608)a,b
Variable
Reported symptom to physician (n ¼ 751) Odds ratio (95% confidence interval)
Sociodemographic characteristics Age (y) 65–74 1.0 75–84 1.07 R85 2.14 Racial background White 1.0 Others 1.07 Highest level of education Less than high school 1.0 (HS)/had HS degree Some college 1.37 Had a college degree 1.46 Had a graduate degree 1.35 Sex Female 1.0 Male 1.25 Annual household income !$15,000 1.0 $15,000 to $24,999 0.92 $25,000 to $34,999 0.86 $35,000 to $49,999 0.82 $50,000 to $74,999 0.85 O75,000 0.72
(0.73–1.57) (0.69–6.63)
(0.53–2.18)
(0.70–2.66) (0.80–2.66) (0.78–2.33)
(0.85–1.83)
(0.41–2.04) (0.40–1.86) (0.38–1.78) (0.37–2.00) (0.30–1.71)
Clinical characteristics Concern beliefs of medicines 1.08 (1.03–1.14) Necessity beliefs of medicines 1.03 (0.98–1.09) Self-rated health Excellent 1.0 Very good 1.69 (0.56–5.09) Good 1.40 (0.47–4.20) Fair 1.54 (0.50–4.80) Poor 4.62 (0.85–24.96) Self-reported adherence Adherent 1.0 Any kind of non-adherence 0.96 (0.66–1.41) Number of medicines taken on a regular basis 1–3 1.0 4–6 1.28 (0.82–2.01) 7–24 1.01 (0.58–1.75) How many times have you seen a physician in the past 6 mo? 1 1.0 2 or more times 4.07 (1.52–10.93) How many times have you visited an emergency room in the past 6 mo? 1 1.0 2 times 1.02 (0.54–1.92) 3 or more times 0.60 (0.17–2.09) (Continued)
231
Table 3 (Continued )
Variable
Reported symptom to physician (n ¼ 751) Odds ratio (95% confidence interval)
How many times have you been admitted in the hospital in the past 6 mo? 1 1.0 2 times 1.05 (0.58–1.92) 3 or more times 2.00 (0.56–7.08) Behavioral characteristics Number of pharmacies 0 1.0 1 1.27 (0.57–3.37) 2 1.71 (0.67–4.40) 3-10 3.60 (1.10–11.13) Do you have to pay any of your prescription drug costs? No prescription insurance 1.0 Yes, pays part of costs 1.36 (0.87–2.02) No, pays none of costs 3.80 (1.57–11.13) Do you think the symptom experienced was due to a medication? Not related to medicine 1.0 Related it to a medicine 2.71 (1.76–4.17) Bolded values indicate significance; *P ! 0.05. a Model contains sociodemographic data, clinical characteristics, and behavioral characteristics, n ¼ 637 represents actual number of respondents included in analysis after missing data. b Pseudo-R2 statistics ¼ 0.174; c2 ¼ 8.89, df ¼ 8, P O 0.1 (Hosmer and Lemeshow test).
Patients who had seen a physician R2 times in the past 6 months were more likely to report symptoms. This makes sense because people who frequently see their physicians have ample opportunity to report symptoms and will report those that are important.7 Also, such patients may be more ill than others yet they might believe that their health is being adequately monitored because of their increased physician visits. Medicare enrollees have insurance and access, in theory, to physicians. However, social barriers, such as transportation, language, and knowledge to navigate the US healthcare system, are likely to affect ability to report symptoms. Respondents who perceived that the symptoms experienced were due to a medication were more likely to report symptoms to their physician than those who did not relate the symptom to a medicine. Individuals may perceive that their medication could be changed or the side effects from the medication relieved if physicians knew about the symptoms. This finding corroborates the notion that attribution to medicines is key to symptom
232
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233
reporting. Thus, appropriate attributions are critical, and healthcare providers need to assess symptomatology. Using more than 3 pharmacies to obtain prescription medication in a typical month was related to reporting symptoms to physicians compared with using no pharmacy. Patients using more pharmacies may likely be using more prescription medications, may be worried about their health, and/or inclined to seek healthcare. Perhaps, their drug plan requires them to use mail order; therefore, creating the possibility of attributing the symptoms experienced to their medicines and subsequently reporting them to their physician. Alternatively, a previous study8 showed that using fewer pharmacies may reduce self-reported ADEs. The rationale is that it may be easier for physicians to obtain a complete list of medications if the individual uses 1 pharmacy. Having multiple medications was not significantly related to reporting symptoms to the physician. This was not consistent with previous literature,7 and this may have occurred because other sociopsychological variables related to medication use were accounted for, such as the concern and necessity beliefs in medicines. These variables may be more important than the number of medicines used by patients, because how symptoms are interpreted and attributed to medications may be based on motivation to tolerate the effects and past experience. Patients who did not have to pay part of their prescription costs were more likely to report symptoms to their physician. This relationship was not expected as individuals who paid part of their prescription costs were thought to be more likely to report symptoms to their physicians. Such individuals would be financially involved in the purchase of their medicines and would therefore attempt to ensure that they were getting the best results by reporting any side effects. However, it is possible that patients who did not pay part of their prescription costs perceived that they could report any symptoms they were experiencing and get a replacement medication easily. They would, therefore, report the symptoms more to the physician because there were no financial consequences or investments to doing that. Alternatively, individuals who paid part of the prescription costs would be more likely to tolerate the symptoms they were experiencing because of their financial investment involved in getting another medication and, therefore, would be less likely to report the symptom to the physician. In addition,
more generous prescription benefits may be related to more generous physician benefits. People who did not pay any part of their prescription costs may have had no cost sharing or minimal cost sharing for physician visits, making it easier to access physicians to report symptoms. This study has limitations. The reporting of symptoms was based on 10 recalled symptoms, and individuals might have experienced other symptoms as well as forgotten if they had experienced a particular symptom. For example, a person may be more likely to remember having symptoms if they caused great concern. Respondents were asked only about reporting symptoms to physicians, and it was not determined whether individuals had experienced these symptoms in the past or previously reported the symptoms to physicians. One might also expect that patients who experience more severe or frequent symptoms would be more likely to report this information to physicians, but no data were gathered about the severity and frequency of symptoms. Also, individual tolerance differences in the intensity of symptoms were not examined. Finally, the subjects used in this study were online users and are not generalizable to all older adults in the United States. They were mostly white and relatively healthy. Thus, the authors anticipate that this study represents a conservative estimate regarding symptom experience among Medicare enrollees. It is unclear whether and how other demographic characteristics would affect attribution to medicines. Assessing beliefs in medications may help us to identify unreported symptoms and subsequent ADEs. Future research should examine how patient’s medication attribution leads to identification and reporting of ADEs. Moreover, symptom reporting to physicians and other health professionals should be encouraged so that preventable ADEs are detected. In conclusion, seeing physicians 2 or more times in the past 6 months, having higher concern beliefs about medicines, perceiving that symptoms experienced were due to medications, using more than 3 pharmacies to obtain prescription medication, and not paying part of the prescription drug costs were related to reporting symptoms to physicians. References 1. Dewitt J, Sorofman B. A model for understanding patient attribution of adverse drug reaction symptoms. Drug Inf J 1999;33(3):907–920. 2. Leventhal H, Nerenz D, Straus A. Self-regulation and the mechanisms for symptom appraisal. In:
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233
3.
4.
5. 6.
7.
8.
9.
Mechanic D, ed. Symptoms, Illness Behavior and Help Seeking. New York: Prodist; 1978. p. 55–86. Leventhal H, Leventhal E, Cameron L. Representations, Procedures and Affect in Illness Self-Regulation: A PerceptualdCognitive Model. In: Baum A, Revenson T, Singer J, eds. Handbook of Health Psychology. New York: Lawrence Erlbaum; 1998. Bates DW, Cullen DJ, Laird N, Peterson LA, Small SD, Servi D, et al. Incidence of adverse drug events and potential adverse drug events: implications for prevention. JAMA 1995;274:29–34. IOM. Preventing Medication Errors. Washington, DC: National Academy Press; 2006. Gandhi T, Weingart S, Borus J, Seger A, Peterson J, Burdick E, et al. Adverse drug events in ambulatory care. N Engl J Med 2003;348:1556–1564. Weingart S, Gandhi T, Seger A, Seger D, Borus J, Burdick E, et al. Patient-reported medication symptoms in primary care. Arch Intern Med 2005;165: 234–240. Oladimeji OO, Farris KB, Urmie JG, Doucette WR. Risk factors for self-reported adverse drug events among Medicare enrollees. Ann Pharmacother 2008; 42:53–61. doi:10.1345/aph.1K073. Diefenbach MA, Levanthal H. The common-sense model of illness representations: theoretical and
10.
11.
12.
13.
14.
15. 16.
233
practical considerations. J Social Distress Homeless 1996;5:11–38. Levanthal H, Diefenbach M, Levanthal EA. Illness cognition: using common sense to understand treatment adherence and affect cognition interactions. Cognit Ther 1992;16:143–163. Aversa S, Kimberlin C, Segal R. The medication attribution scale: perceived effects of anti-retrovirals and quality of life. Qual Life Res 1998;7(3):205–214. Horne R, Weinman J, Hankins M. The beliefs about medicines questionnaire: the development and evaluation of a new method for assessing the cognitive representation of medication. Psych Health 1999; 14:1–24. Horne R, Weinman J. Patients beliefs about prescribed medicines and their role in adherence to treatment in chronic physical illness. J Psychosom Res 1999;47(6):555–567. Idler E, Benyamini Y. Self-rated health and mortality: a review of twenty-seven community studies. J Health Soc Behav 1997;38:21–37. Bailis D, Segall A, Chipperfield J. Two views of selfrated health status. Soc Sci Med 2003;56(2):203–217. Morisky D, Green L, Levine D. Concurrent and predictive validity of a self-reported measure of medication adherence. Med Care 1986;24(1):67–74.
Research in Social and Administrative Pharmacy 5 (2009) 225–233
Original Research
Symptomatology, attribution to medicines, and symptom reporting among Medicare enrollees Olayinka Oladimeji, B.Pharm., Karen B. Farris, B.S.Pharm., Ph.D.*, Julie G. Urmie, Ph.D., William R. Doucette, Ph.D. Department of Pharmaceutical Socioeconomics, University of Iowa, Iowa City, IA 52242, USA
Abstract Background: Adverse drug events (ADEs) are generally preceded by symptoms reported by patients, and understanding who reports symptoms is important in understanding ADEs. Objectives: The objectives of this study were to (1) determine the prevalence of symptomatology experienced, recalled, and reported to physicians (2) quantify the extent of attribution of recalled symptoms to medicines, and (3) predict factors associated with the reporting of recalled symptoms. Methods: This was an Internet-based survey administered by Harris InteractiveÒ. Data collected and included in the analysis were symptoms experienced, symptom attribution to medicines, number of medicines, concern and necessity beliefs in medicines, self-rated health, number of physician visits, hospitalizations and emergency room visits in the past 6 months, whether patients had to pay part of their prescription costs, number of pharmacies, self-reported medication adherence, and sociodemographics. A descriptive analysis of sociodemographic and clinical/behavioral characteristics was completed, the frequency distribution of symptoms and their attribution to medication was obtained, and independent factors that predicted who was likely to report symptoms to physicians, using a multiple logistic regression analysis, were identified. Results: Sixty-two percent (n ¼ 751) reported having at least 1 of 10 symptoms in the past month. Of the 1220 Medicare enrollees, 6.0-31.1% experienced varied specific symptoms. Of those who recalled they had experienced a particular symptom, 11.7-48.8% thought the symptoms were related to the medicine they were taking. Reporting symptoms to physicians was positively related to concern beliefs, seeing physicians R 2 times in the past 6 months, perceiving symptom experienced was due to medicines, using more than 3 pharmacies, and not having to pay part of prescription costs. Conclusions: Assessing beliefs in medications may help to identify unreported symptoms and subsequent ADEs. In addition, symptom reporting to physicians and other health professionals should be encouraged so that preventable ADEs are detected. Ó 2009 Elsevier Inc. All rights reserved. Keywords: Symptom reporting; Adverse drug events; Medicare
* Corresponding author. Division of Clinical and Administrative Pharmacy, S-525 Pharmacy Building, 115 S. Grand Avenue, Iowa City, IA 52242-1112, USA. Tel.: þ1 319 384 4516; fax: þ1 319 353 5646. E-mail address: [email protected] (K.B. Farris). 1551-7411/09/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.sapharm.2008.08.004
226
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233
Introduction Adverse drug events (ADEs) are generally preceded by symptoms reported by patients, and the process by which symptoms are labeled may be applicable to the attribution of ADEs.1 Symptoms serve as triggers for cognitive processes, and the ability of patients to identify a symptom as possibly drug-induced is tied closely to the process by which patients label them.2,3 Symptoms are, therefore, the basis for identifying an ADE but not the definition of it, which is an injury resulting from medical interventions related to the use of a drug.4,5 Patients may have symptoms for months without any changes in their medications, and only a small percentage of patients report that symptoms actually led to a physician visit.6 However, Weingart et al reported that patients discussed 196 of their 286 medication symptoms with their physicians and 22% of the symptoms were serious enough to require a visit to a medical facility. Twenty-three percent of patients’ unreported medication symptoms led to preventable or ameliorable ADEs. Overall, 65% of the cases had ameliorable ADEs, which occurred as a result of failure of physicians to act on self-reported symptoms. They also reported that for every symptom that patients experienced but failed to report, 1 in 5 resulted in an ADE that could have been prevented or been made less severe. They did not inquire into patient’s reasons for not reporting some of the adverse side effects they experienced. It was speculated that patients may have been reticent about discussing side effects that were either common or relatively mild or that were embarrassing such as diarrhea or sexual dysfunction. Thus, patient-reported medication symptoms represent a valuable source of information about medication safety.7 Various factors have been identified in the reporting of symptoms. Patients who take more medications and have more drug allergies are more likely to report medication symptoms to their physicians. These patients may also have more physician encounters and, therefore, more opportunities to report symptoms.7 Patients with fewer symptoms are probably able to manage and tolerate them, whereas those with more symptoms would rather seek the help of a health professional. Patients who report more symptoms are also more likely to report an ADE to their physicians.8 The common sense model of illness is a framework that attempts to explain how people may adapt to and/or manage health threats such as
symptoms. This model may be useful to examine the representation of symptoms experienced by Medicare enrollees and their coping strategy for dealing with it, reporting symptoms to physicians. The common sense model of illness views individuals as active problem solvers who strive to assign meaning to a somatic sensation, such as symptoms, and this typically involves a representation of the symptom and a coping procedure for dealing with the symptom.9 There are typically 5 distinct attributes of illness representations that may be evoked when a person experiences a symptom, including identity, timeline, cause, control, and consequences. With reference to reporting of symptoms and their attribution to medications and to ADEs, individuals may consider all of these attributes.9-11 In terms of measuring attribution to medications, perceived attribution of a symptom to a medication has been assessed using a constructed scale, the medication attribution scale (MAS). Aversa et al found that negative attributions about the effects of anti-retroviral medications were related to increased depressive symptoms, discontinuation of therapy, and quality of life.11 Weingart et al measured perceived medication attribution with a single item.7 Horne et al used the common sense model of illness to show that necessity and concern beliefs about medicines are 2 core themes that people use when interpreting symptoms and causal attributions related to their medications.12,13 Necessity beliefs relate to an individual’s perception of the necessity of medication for maintaining health, whereas concern beliefs relate to an individual’s concern about the adverse consequences of medication based on beliefs about the potential for dependence or harmful long-term effect. Items from the necessity scale include, for example, ‘‘My health at present depends on my medicines’’ and ‘‘Without medicines, I would be very ill.’’ Concern beliefs comprise items such as ‘‘I sometimes worry about the long-term effects of my medicines,’’ ‘‘Having to take my medicine worries me,’’ and ‘‘My medicines disrupt my life.’’ Significant relationships between medication beliefs and adherence have been reported, with higher scores on the concern-belief scale correlating with lower reported adherence. In addition, higher reported adherence rates were associated with higher necessity-concerns difference scores. The concern-belief scale, therefore, assesses the negative attitude of individual patients toward medication, whereas the necessity-belief scale assesses their positive attitude toward medication.12,13
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233
It is hypothesized that patients with higher scores on the concern-beliefs scale may be more cognitively aware of their symptoms and pay particular attention to unwanted reactions and are, therefore, likely to report symptoms. Higher scores on the scale indicate stronger negative attitudes toward taking medication, not positive beliefs, because the individuals worry more about the long-term effects of their medicines and their dependency on their medicines. In addition, it is hypothesized that patients with higher scores on the necessity-beliefs scale would be more likely to report symptoms because they are more likely to take their medications and may monitor the effects, making them more aware of reactions. At the same time, the opposite could occur for necessity beliefs. For example, patients could be more tolerant of side effects because they value the necessity of continuing their therapy and may be less likely to report their symptoms. The objectives of this study were to (1) determine the prevalence of symptomatology experienced, recalled, and reported to physicians (2) quantify the extent of attribution of recalled symptoms to medicines, and (3) predict factors associated with the reporting of recalled symptoms.
Methods Design An Internet-based survey was administered by Harris InteractiveÒ on behalf of the University of Iowa, College of Pharmacy. University of Iowa investigators designed the survey and analyzed the data. The project was approved by the Institutional Review Board of the University of Iowa. Patients/setting Harris InteractiveÒ maintains a confidential panel of individuals who have agreed to be invited to participate in telephone and/or online surveys. Harris InteractiveÒ invited individuals in their online panel to participate in this study, and participants received credit from Harris InteractiveÒ for completing the online survey. The inclusion criteria were being 65 years or older, English speakers, US residents, and enrolled in Medicare. Harris InteractiveÒ used their online panel of potential subjects and provided data to University of Iowa researchers from a non-probability sample of 1220 anonymous respondents.
227
Data collection Respondents completed a 166-item survey, which was administered between October 13 and October 19, 2005. The survey took approximately 18 minutes to complete, as numerous skip patterns were included for questions that did not apply to some respondents. In addition, respondents could answer part of the survey and return later if necessary for completion. The survey was pilot tested in Iowa City using a convenience sample of 30 older adults at the local senior citizens center. These individuals completed the survey and provided comments about the clarity of survey questions. Also, students and colleagues completed the online version before its release. Information on health symptoms experienced and recalled in the past month (yes/no) was collected. The ‘‘past month’’ time frame was used to improve recall. The symptoms measured were headaches, dizziness, or problems with balance; stomach or gastrointestinal problems; muscle aches, incontinence, or problems with urinating; rash or itching; problems with sleep; changes in mood, fatigue, and sexual problems. The predetermined list of symptoms was based on a previous study that identified ADEs.7 There was no opportunity for respondents to report other nonlisted symptoms. Respondents were also asked if they thought the symptoms they experienced were related to a medicine they were taking.7 Next, they were asked if they reported the symptom to their physician. If no, they were asked if they intended to report the symptom to their physician in the future. Data used in this analysis included age, race, sex, education, income, number of medicines used on a regular basis, concern and necessity beliefs in medicines, self-rated health, number of physician visits in the past 6 months, number of hospitalizations in the past 6 months, number of emergency room visits in the past 6 months, whether they had to pay part of their prescription costs, number of pharmacies where they got their prescription medicines, and self-reported medication adherence. For sociodemographic data, the age of respondents, racial background, sex, and the highest level of education completed were determined. Respondents were asked about their self-rated health status. Respondents rated their health compared with other individuals of their own age, and a 5-item response scale anchored with poor and excellent was used.14,15 To determine the number of medications used, the number of different prescription medicines
228
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233
used in the past month was asked. Then, respondents indicated the number of medications they took on a regular basis, among those they had taken in the past month. To examine their concerns and necessity beliefs about their medications, the 10 items from Horne et al scale were used.11 Five items ask about concern beliefs and include items such as ‘‘Having to take medicines worries me’’ and ‘‘I sometimes worry about the long-term effects of my medicines.’’ Necessity beliefs were assessed with 5 items, including, for example, ‘‘My life would be impossible without my medicines’’ and ‘‘Without medicines, I would be very ill.’’ Five-point Likert scales anchored with strongly disagree and strongly agree were used. Items for concern and necessity beliefs were summed separately, providing a scale with a range of 5-25 for either scale. Previous studies using these scales show reliability estimates ranging from 0.65 to 0.86 and its construct validity was established using factor analysis and hypothesis testing.11 The respondents stated the number of times they had seen a physician, visited the hospital, or been admitted to the emergency room in the past 6 months. They also indicated whether they had to pay part of their prescription drug costs. Respondents indicated the number of pharmacies where they had obtained their prescription medicines in a typical month. Medication adherence was measured using 4 items from the Morisky scale.16 Questions such as ‘‘During the past month, have you ever forgotten to take your medication?’’ and ‘‘When you feel better, do you sometimes stop taking your medication?’’ were answered by the respondents. These 4 items were summed in the analyses. Analysis All analyses were completed by investigators at the University of Iowa. A descriptive analysis of sociodemographic and clinical/behavioral characteristics was completed (Table 1). The frequency distribution of symptoms, respondents’ attribution to medication, and the percentage of those who reported each symptom to physicians were determined (Table 2). Independent factors that predicted who was more likely to report symptoms to physicians, using a multiple logistic regression analysis to obtain odds ratios (ORs) with 95% confidence intervals, were identified (Table 3). The dependent variable was reporting at least 1 symptom to
physicians, among individuals who experienced any symptom in the past month. Sociodemographic data, clinical characteristics, and behavioral characteristics were used as independent variables in the regression. Most independent variables were recoded into dummy variables. Respondents who had not seen a physician in the past 6 months were excluded because they were unlikely to be able to report symptoms to physicians in the ‘‘past month.’’ Also, 6 months was thought to be an appropriate recall period for utilization of health services. Those who had missing data from other variables were also excluded from the analysis. Statistical analyses were performed using SPSS software (version 15.0). Results Participants were between 65 and 94 years old, and 54.9% were female. Most respondents were white, had some college experience, used more than 1 prescription medicine on a regular basis, obtained their prescription medicines from 1 pharmacy, and perceived themselves to be relatively in good health (Table 1). Sixty-two percent (n ¼ 751) reported having at least 1 of the 10 symptoms in the past month. Of the 1220 Medicare enrollees, a range of 6.0-31.1% experienced a particular symptom (Table 2). Of those who stated they had experienced a particular symptom, 11.7-48.8% thought the symptoms were related to the medicine they were taking. A range of 34.3-55.2% reported the particular symptoms to their physicians, depending on the symptom. Consistent with previous literature, patients reported symptoms related to muscular aches, fatigue, sleep problems, sexual problems, and gastrointestinal problems (43.2%, 48.1%, 44.4%, 48.8%, and 50.4%, respectively) more often to their physicians than headache and incontinence (34.3% and 41.0%).5 Patients were least likely to report headaches. Sixty percent of individuals having at least 1 of the 10 symptoms in the past month reported 1 or more of their symptoms to their physicians. About two thirds of individuals with an unreported symptom had an intention to report at least 1 symptom. In the full model, reporting symptoms to physicians was related to concern beliefs about medications (OR ¼ 1.08, 95% CI ¼ 1.03-1.14), seeing physicians R2 times in the past 6 months (OR ¼ 4.07, 95% CI ¼ 1.5210.93), perceiving symptom experienced was due to a medicine (OR ¼ 2.71, 95% CI ¼ 1.76-4.17),
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233 Table 1 Descriptive characteristics of the study populationa (n¼1220)
Variable
Number (%)
Sociodemographics Age (y), mean 65–74 75–84 R85
746 (61.1) 441 (36.1) 31 (2.5)
Sex Male Female
543 (45.1) 661 (54.9)
Racial background White Hispanic African American Other
Mean Standard deviation 72.89 5.68
1106 59 18 25
(90.7) (4.8) (1.5) (2.3)
Highest level of education Less than high school (HS)/had HS degree Some college Had a college degree Had a graduate degree
420 (34.5) 199 (16.3) 209 (17.1)
Annual household income Less than $15,000 $15,000 to $24,999 $25,000 to $34,999 $35,000 to $49,999 $50,000 to $74,999 O75,000
81 196 262 270 191 149
199 (16.3)
(6.6) (16.1) (21.5) (22.1) (15.7) (12.2)
Clinical characteristics Self-rated health Excellent Very good Good Fair Poor Number of medicines 0 1–3 4–6 7–24 Self-reported adherence 0 1 2
151 423 439 184 20
(12.4) (34.8) (36.1) (15.1) (1.6)
133 456 284 257
(10.9) (37.4) (23.3) (21.1)
Table 1 (Continued )
Variable
Number (%)
592 (48.5) 286 (23.4) 152 (12.5) (Continued)
Mean Standard deviation
3 26 (2.1) 4 5 (0.4) Concern beliefs of 11.96 3.97 medicines (ranges from 5-25), mean SD (lower scores is less concern) Necessity beliefs of 16.62 4.81 medicines (ranges from 5-25, mean SD (higher scores is more beliefs) Times seen a physician in the past 6 mo 0 116 (9.6) 1 302 (24.9) 2 400 (33.0) 3 or more times 394 (32.5) Times visited the emergency department in the past 6 mo 0 1054 (86.8) 1 132 (10.9) 2 23 (1.9) 3 or more times 6 (0.5) Times admitted in the hospital in the past 6 mo 0 1040 (85.2) 1 147 (12.1) 2 24 (2.0) 3 or more times 5 (0.4) Behavioral characteristics Do you have to pay part of Yes, pays part of costs No, pays none of costs Has no prescription insurance Number of pharmacies 0 1 2 3–10
Geographic region (state of residence)b North East 257 (20.9) Midwest 393 (32.2) South 319 (26.1) West 248 (20.4)
229
your prescription costs? 467 (62.2) 42(5.6) 224 (29.8)
127 852 201 39
(10.4) (69.9) (16.5) (3.2)
a Values are number (percentage) except otherwise indicated. Numbers that do not add up to 1220 indicate missing data. b The North East region includes Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, and Vermont. Midwest region includes Iowa, Indiana, Illinois, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, and Wisconsin. South region includes Alabama, Arkansas, Delaware, District of Columbia, Florida, Georgia, Kentucky, Louisiana, Maryland, North Carolina, Oklahoma, South Carolina, Tennessee, Texas, Virginia, and West Virginia. West region includes Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, and Wyoming.
230
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233
Table 2 Self-reported symptoms in the past month among medicare enrollees (n ¼ 1220)a Symptoms
Experienced any symptom n (%)a
Related it to medicine n (%)b
Reported symptom to physician n (%)b
Muscle aches Sleep problems Fatigue Stomach problems Dizziness Incontinence Headache Rash itching Sexual problems Mood changes
380 367 337 234 192 195 172 140 125 73
56 43 75 60 66 23 24 40 60 24
164 163 162 118 106 80 59 71 61 30
a b
(31.1) (30.1) (27.6) (19.2) (15.7) (15.8) (14.1) (11.5) (10.2) (6.0)
(14.7) (11.7) (22.2) (25.6) (34.4) (11.8) (24.0) (28.6) (48.0) (32.9)
(43.2) (44.4) (48.1) (50.4) (55.2) (41.0) (34.3) (50.7) (48.8) (41.1)
Denominator in column 2 is 1220. Denominator in columns 3 and 4 is the frequency of experiencing that specific symptom (number shown in column 2).
using more than 3 pharmacies (OR ¼ 3.60, 95% CI ¼ 1.10-11.13), and not having to pay part of the prescription costs (OR ¼ 3.80, 95% CI ¼ 1.5711.13) (Table 3). Discussion Understanding who reports symptoms may help us understand ADEs and their reporting. In this study, 62% of individuals reported having at least 1 symptom in the past month and a range of 6.0-31.0% experienced a particular symptom. Concern beliefs in medicines, having 2 or more physician visits, using more than 3 pharmacies, and perceiving symptom experienced was due to medicines were associated with reporting symptoms to physicians. Necessity beliefs about medicines were not related to symptom reporting. The attributes of symptoms and the resulting coping strategies to manage symptoms should align.9 For example, if a patient experiencing a symptom can label, attribute its cause to a medicine rather than age or a medical condition, and believe cure/control of the symptom can be achieved by seeking the help of a health provider, the patient may be likely to report the symptom to a health provider. The results showed that over half the respondents experienced at least 1 symptom in the past month; however, the attribution to medications did not seem high. Aversa et al, using the MAS, showed that about 39-52% respondents attributed the symptoms they were experiencing to the medicine they were taking.11 Also, in a previous study, it was found that about 18% of elderly patients reported an unwanted reaction to the medicine they were taking.8 From these results, it can be inferred that some of the
attributions to medications were lower than expected. Lower than expected attribution rates may have occurred because symptoms were not caused by medications. Respondents may also have attributed symptoms to other causes such as disease or age, or respondents may have lacked sufficient information and/or experience to attribute the symptom to medications. Consistent with the findings from Weingart et al., it is important for health providers to verify symptoms that patients may perceive as normal or due to aging to be sure they are not related to the medicine they are taking.7 As hypothesized, stronger concern beliefs about medicines were positively related to reporting symptoms to physicians, although the ORs were modest. People with stronger concern beliefs might be more sensitive to their symptoms and more watchful for unwanted reactions and thus more likely to report symptoms. Contrary to the hypothesis, necessity beliefs in medicines were not significantly related to symptom reporting. Necessity beliefs in medicines are important in adherence but may not be important in symptom reporting because individuals believe that they are dependent on their medicines and, therefore, may be more willing to tolerate symptoms that do arise and thus not report symptoms to their physicians. Respondents were not asked if they had experienced the particular symptoms in the past. If so, respondents may have learned previous coping strategies for managing symptoms. An important issue with these beliefs is that they may not be generalizable to all patients’ medications, and this inconsistency in specificity may have contributed to the low OR for concern beliefs and the nonsignificant finding for necessity beliefs.
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233 Table 3 Logistic regression predicting reporting of symptoms to physicians (n ¼ 608)a,b
Variable
Reported symptom to physician (n ¼ 751) Odds ratio (95% confidence interval)
Sociodemographic characteristics Age (y) 65–74 1.0 75–84 1.07 R85 2.14 Racial background White 1.0 Others 1.07 Highest level of education Less than high school 1.0 (HS)/had HS degree Some college 1.37 Had a college degree 1.46 Had a graduate degree 1.35 Sex Female 1.0 Male 1.25 Annual household income !$15,000 1.0 $15,000 to $24,999 0.92 $25,000 to $34,999 0.86 $35,000 to $49,999 0.82 $50,000 to $74,999 0.85 O75,000 0.72
(0.73–1.57) (0.69–6.63)
(0.53–2.18)
(0.70–2.66) (0.80–2.66) (0.78–2.33)
(0.85–1.83)
(0.41–2.04) (0.40–1.86) (0.38–1.78) (0.37–2.00) (0.30–1.71)
Clinical characteristics Concern beliefs of medicines 1.08 (1.03–1.14) Necessity beliefs of medicines 1.03 (0.98–1.09) Self-rated health Excellent 1.0 Very good 1.69 (0.56–5.09) Good 1.40 (0.47–4.20) Fair 1.54 (0.50–4.80) Poor 4.62 (0.85–24.96) Self-reported adherence Adherent 1.0 Any kind of non-adherence 0.96 (0.66–1.41) Number of medicines taken on a regular basis 1–3 1.0 4–6 1.28 (0.82–2.01) 7–24 1.01 (0.58–1.75) How many times have you seen a physician in the past 6 mo? 1 1.0 2 or more times 4.07 (1.52–10.93) How many times have you visited an emergency room in the past 6 mo? 1 1.0 2 times 1.02 (0.54–1.92) 3 or more times 0.60 (0.17–2.09) (Continued)
231
Table 3 (Continued )
Variable
Reported symptom to physician (n ¼ 751) Odds ratio (95% confidence interval)
How many times have you been admitted in the hospital in the past 6 mo? 1 1.0 2 times 1.05 (0.58–1.92) 3 or more times 2.00 (0.56–7.08) Behavioral characteristics Number of pharmacies 0 1.0 1 1.27 (0.57–3.37) 2 1.71 (0.67–4.40) 3-10 3.60 (1.10–11.13) Do you have to pay any of your prescription drug costs? No prescription insurance 1.0 Yes, pays part of costs 1.36 (0.87–2.02) No, pays none of costs 3.80 (1.57–11.13) Do you think the symptom experienced was due to a medication? Not related to medicine 1.0 Related it to a medicine 2.71 (1.76–4.17) Bolded values indicate significance; *P ! 0.05. a Model contains sociodemographic data, clinical characteristics, and behavioral characteristics, n ¼ 637 represents actual number of respondents included in analysis after missing data. b Pseudo-R2 statistics ¼ 0.174; c2 ¼ 8.89, df ¼ 8, P O 0.1 (Hosmer and Lemeshow test).
Patients who had seen a physician R2 times in the past 6 months were more likely to report symptoms. This makes sense because people who frequently see their physicians have ample opportunity to report symptoms and will report those that are important.7 Also, such patients may be more ill than others yet they might believe that their health is being adequately monitored because of their increased physician visits. Medicare enrollees have insurance and access, in theory, to physicians. However, social barriers, such as transportation, language, and knowledge to navigate the US healthcare system, are likely to affect ability to report symptoms. Respondents who perceived that the symptoms experienced were due to a medication were more likely to report symptoms to their physician than those who did not relate the symptom to a medicine. Individuals may perceive that their medication could be changed or the side effects from the medication relieved if physicians knew about the symptoms. This finding corroborates the notion that attribution to medicines is key to symptom
232
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233
reporting. Thus, appropriate attributions are critical, and healthcare providers need to assess symptomatology. Using more than 3 pharmacies to obtain prescription medication in a typical month was related to reporting symptoms to physicians compared with using no pharmacy. Patients using more pharmacies may likely be using more prescription medications, may be worried about their health, and/or inclined to seek healthcare. Perhaps, their drug plan requires them to use mail order; therefore, creating the possibility of attributing the symptoms experienced to their medicines and subsequently reporting them to their physician. Alternatively, a previous study8 showed that using fewer pharmacies may reduce self-reported ADEs. The rationale is that it may be easier for physicians to obtain a complete list of medications if the individual uses 1 pharmacy. Having multiple medications was not significantly related to reporting symptoms to the physician. This was not consistent with previous literature,7 and this may have occurred because other sociopsychological variables related to medication use were accounted for, such as the concern and necessity beliefs in medicines. These variables may be more important than the number of medicines used by patients, because how symptoms are interpreted and attributed to medications may be based on motivation to tolerate the effects and past experience. Patients who did not have to pay part of their prescription costs were more likely to report symptoms to their physician. This relationship was not expected as individuals who paid part of their prescription costs were thought to be more likely to report symptoms to their physicians. Such individuals would be financially involved in the purchase of their medicines and would therefore attempt to ensure that they were getting the best results by reporting any side effects. However, it is possible that patients who did not pay part of their prescription costs perceived that they could report any symptoms they were experiencing and get a replacement medication easily. They would, therefore, report the symptoms more to the physician because there were no financial consequences or investments to doing that. Alternatively, individuals who paid part of the prescription costs would be more likely to tolerate the symptoms they were experiencing because of their financial investment involved in getting another medication and, therefore, would be less likely to report the symptom to the physician. In addition,
more generous prescription benefits may be related to more generous physician benefits. People who did not pay any part of their prescription costs may have had no cost sharing or minimal cost sharing for physician visits, making it easier to access physicians to report symptoms. This study has limitations. The reporting of symptoms was based on 10 recalled symptoms, and individuals might have experienced other symptoms as well as forgotten if they had experienced a particular symptom. For example, a person may be more likely to remember having symptoms if they caused great concern. Respondents were asked only about reporting symptoms to physicians, and it was not determined whether individuals had experienced these symptoms in the past or previously reported the symptoms to physicians. One might also expect that patients who experience more severe or frequent symptoms would be more likely to report this information to physicians, but no data were gathered about the severity and frequency of symptoms. Also, individual tolerance differences in the intensity of symptoms were not examined. Finally, the subjects used in this study were online users and are not generalizable to all older adults in the United States. They were mostly white and relatively healthy. Thus, the authors anticipate that this study represents a conservative estimate regarding symptom experience among Medicare enrollees. It is unclear whether and how other demographic characteristics would affect attribution to medicines. Assessing beliefs in medications may help us to identify unreported symptoms and subsequent ADEs. Future research should examine how patient’s medication attribution leads to identification and reporting of ADEs. Moreover, symptom reporting to physicians and other health professionals should be encouraged so that preventable ADEs are detected. In conclusion, seeing physicians 2 or more times in the past 6 months, having higher concern beliefs about medicines, perceiving that symptoms experienced were due to medications, using more than 3 pharmacies to obtain prescription medication, and not paying part of the prescription drug costs were related to reporting symptoms to physicians. References 1. Dewitt J, Sorofman B. A model for understanding patient attribution of adverse drug reaction symptoms. Drug Inf J 1999;33(3):907–920. 2. Leventhal H, Nerenz D, Straus A. Self-regulation and the mechanisms for symptom appraisal. In:
Oladimeji et al./Research in Social and Administrative Pharmacy 5 (2009) 225–233
3.
4.
5. 6.
7.
8.
9.
Mechanic D, ed. Symptoms, Illness Behavior and Help Seeking. New York: Prodist; 1978. p. 55–86. Leventhal H, Leventhal E, Cameron L. Representations, Procedures and Affect in Illness Self-Regulation: A PerceptualdCognitive Model. In: Baum A, Revenson T, Singer J, eds. Handbook of Health Psychology. New York: Lawrence Erlbaum; 1998. Bates DW, Cullen DJ, Laird N, Peterson LA, Small SD, Servi D, et al. Incidence of adverse drug events and potential adverse drug events: implications for prevention. JAMA 1995;274:29–34. IOM. Preventing Medication Errors. Washington, DC: National Academy Press; 2006. Gandhi T, Weingart S, Borus J, Seger A, Peterson J, Burdick E, et al. Adverse drug events in ambulatory care. N Engl J Med 2003;348:1556–1564. Weingart S, Gandhi T, Seger A, Seger D, Borus J, Burdick E, et al. Patient-reported medication symptoms in primary care. Arch Intern Med 2005;165: 234–240. Oladimeji OO, Farris KB, Urmie JG, Doucette WR. Risk factors for self-reported adverse drug events among Medicare enrollees. Ann Pharmacother 2008; 42:53–61. doi:10.1345/aph.1K073. Diefenbach MA, Levanthal H. The common-sense model of illness representations: theoretical and
10.
11.
12.
13.
14.
15. 16.
233
practical considerations. J Social Distress Homeless 1996;5:11–38. Levanthal H, Diefenbach M, Levanthal EA. Illness cognition: using common sense to understand treatment adherence and affect cognition interactions. Cognit Ther 1992;16:143–163. Aversa S, Kimberlin C, Segal R. The medication attribution scale: perceived effects of anti-retrovirals and quality of life. Qual Life Res 1998;7(3):205–214. Horne R, Weinman J, Hankins M. The beliefs about medicines questionnaire: the development and evaluation of a new method for assessing the cognitive representation of medication. Psych Health 1999; 14:1–24. Horne R, Weinman J. Patients beliefs about prescribed medicines and their role in adherence to treatment in chronic physical illness. J Psychosom Res 1999;47(6):555–567. Idler E, Benyamini Y. Self-rated health and mortality: a review of twenty-seven community studies. J Health Soc Behav 1997;38:21–37. Bailis D, Segall A, Chipperfield J. Two views of selfrated health status. Soc Sci Med 2003;56(2):203–217. Morisky D, Green L, Levine D. Concurrent and predictive validity of a self-reported measure of medication adherence. Med Care 1986;24(1):67–74.