symptom perception and adherence to asthma controller medications

Clinical Scholarship

Symptom Perception and Adherence to AsthmaController MedicationsRuth Ohm, Lauren S. Aaronson

Purpose: To explore asthma symptom perception and the relationship between asthma symp-tom perception and adherence to asthma treatment.

Design: Adult patients (N=120) of asthma/allergy specialty clinics, taking Advair� as a con-troller medication, were enrolled in this cross-sectional descriptive study.

Methods: Ninety-seven participants completed 4 weeks of daily diaries to assess subjectivesymptom perception and measured peak expiratory flow rates (PEFR), both done twicedaily. Individual perceptual accuracy scores (PAS) were determined by correlating the sub-jective symptom perception scores with the PEFRs. Measures included demographic vari-ables, illness identity (personal control and treatment control, consequences, and timeline-cyclical subscales of the IPQ-R), asthma severity (FEV1 percentage) and a single-item indi-cator of perceived asthma severity. Adherence was measured by the Medication AdherenceReport Scale (MARS) and by an Advair� dose count (percentage of doses taken as pre-scribed).

Findings: Independent t tests comparing adherence rates of good versus poor perceiverswere not significant, using either the percentage Advair� dose count or the MARS. Multipleregression analyses showed that years with asthma, illness identity, and peak flow variabilitywere all significant explanatory variables for perceptual accuracy.

Conclusion: Peak flow variability adds complexity to the relationship between perceptualaccuracy and adherence that warrants further investigation.

JOURNAL OF NURSING SCHOLARSHIP, 2006; 38:3, 292-297. C© 2006 SIGMA THETA TAU INTERNATIONAL.

[Key words: symptom perception, adherence, asthma]

* * *

Over 20.3 million people in the United States haveasthma (National Center for Health Statistics,2003) and the associated costs are staggering, es-

timated to be over $16 billion in 2004 (American LungAssociation, 2005). Appropriate management of asthma,including patient education to promote self-management(Centers for Disease Control, 1996; National Asthma Edu-cation and Prevention Program (NAEPP), 1997; 2002), notonly might reduce the frequency of emergency room visitsand hospitalizations for asthma, and hence the costs asso-ciated with asthma, but also should improve the quality oflife for people with asthma.

Because medications are required to control asthma,as well as for rescue from acute attacks, adherence todaily medication regimens and adjustment of medicationsto changes in lung function are necessary for positive healthoutcomes. Difficulty with accurately perceiving decreasedlung function, however, might prevent timely adjustmentsin medications and seeking healthcare intervention. Whilesome research has shown a relationship between poor symp-tom perception and poor outcomes (Banzett, Dempsey,

O’Donnell, & Wamboldt, 2000; Magadle, Berar-Yaney, &Weiner, 2002; Yoos & McMullen, 1999), studies document-ing an association between adherence to therapeutic regi-men and the ability to perceive asthma symptoms are lim-ited (Lehrer, Feldman, Giardino, Song, & Schmaling, 2002;Wamboldt, 1998). Therefore, the purpose of this study wasto explore asthma symptom perception and the relation-ship between asthma symptom perception and adherenceto asthma treatment, as a basis for designing individually-tailored interventions for people with asthma.

Ruth Ohm, RN, PhD, ARNP, Eta Kappa-At-Large, Associate Professor,Baker University School of Nursing, Stormont-Vail HealthCare, Topeka,KS; Lauren S. Aaronson, RN, PhD, FAAN, Delta, Senior Advisor, NationalInstitute of Nursing Research, Professor, University of Kansas School ofNursing, Kansas City, KS. The authors acknowledge help from KU NursesAlumni Research Award, Sigma Theta Tau International, Eta Kappa Chapter-at-Large Research Grant, AstraZeneca and GlaxoSmithKline provided ofpeak flow meters. Correspondence to Dr. Ohm, 3006 SW StaffordshireRoad, Topeka, KS 66614. E-mail: [email protected]

Accepted for publication February 16, 2006.

292 Third Quarter 2006 Journal of Nursing Scholarship

Symptom Perception

Background

Asthma is a chronic inflammatory airway disease char-acterized by episodes of reversible airway obstruction. In-flammation leads to hyperresponsiveness of the airways toa variety of stimuli, including irritants, viral or bacterial in-fections, allergens, and cold air (NAEPP, 1997). Daily use ofan anti-inflammatory medication (such as a corticosteroid)is recommended for all patients with mild, moderate, or se-vere persistent asthma to prevent exacerbations.

Asthma Symptom PerceptionResults from several studies have indicated that a pa-

tient’s ability to perceive decreased lung function, as mea-sured by forced expiratory volume in one second (FEV1),varies markedly. Approximately 15% of patients withasthma experience no sensation of breathlessness or dys-pnea, despite substantial obstruction as measured by FEV1

(Banzett et al., 2000). Possibly these poor perceivers un-derestimate the severity of their asthma and fail to adhereto their therapeutic regimen, predisposing them to air-way hyperreactivity and obstruction (Lehrer et al., 2002;Wamboldt, 1998).

The ability to perceive dyspnea has been measuredby methacholine-histamine challenge, by external resistanceloads, and by correlation of subjective data with peak expi-ratory flow rate (PEFR). Methacholine and histamine provo-cation is used to reduce the FEV1 by 20% or 30%. At thisreduced airflow, symptoms of dyspnea are typically detected.Julius and colleagues (2002), however, found a 40% reduc-tion of FEV1 was required for some to detect their dyspnea.Further, while methacholine and histamine challenges pro-duce alterations of intrinsic respiratory loads, they also areinvasive and unpleasant. Using external respiratory loads,which requires breathing through a mouthpiece with noseclamped, with varying loads applied to the apparatus, is notas invasive or unpleasant, but it requires a laboratory settingwhere the participants’ focus is directed on their respiratorystatus (Harver & Mahler, 1998). The demand characteris-tics of the laboratory setting raise concerns about externalvalidity.

On the other hand, correlating peak expiratory flowrates (PEFR) with subjective ratings of dyspnea removes datacollection from laboratory settings to natural settings wherepeople experience symptoms as part of their daily activities.The PEFR is the highest expiratory flow with maximumforced effort following full inflation of the lungs, typicallyobtained within the first 100 milliseconds of forced expi-ration (Jain & Kavuru, 2000). The peak flow meter (usedto obtain the PEFR) is an inexpensive hand-held instrumentuseful in tracking trends or changes in airway obstructionand is recommended for regular use by people with asthma.For all these reasons, we chose this measure of symptomperception for this study.

Symptom Interpretation ModelThe Symptom Interpretation Model (SIM), developed

to describe how symptoms are perceived, interpreted and

acted upon by patients (Teel, Meek, McNamara, & Watson,1997) was used in this study. The SIM has three key con-structs: input, interpretation, and outcome. Input, or symp-tom awareness, is the recognition of a change or disturbancein the person (i.e., the stimulus must reach a threshold forperception.) Interpretation captures the person’s illness iden-tity and is a cognitive process to compare a stimulus withwhat the person knows or has previously experienced andto appraise the threat level and consequent need for action.Outcome is the response to the symptom, either action orinaction (Teel et al.).

Several factors constitute illness identity, central to In-terpretation in the SIM. These include symptoms associatedwith the illness, cause of the disease, whether it is an episodicor chronic condition, beliefs about severity and long-termeffects on functioning, how amenable the illness is to con-trol or cure (Lau, Bernard, & Hartman, 1989; Leventhal& Nerenz, 1985), and emotional representation and illnesscoherence or overall comprehension of the illness (Moss-Morris et al., 2002).

The primary research hypothesis for the present studywas that accurate, or good, symptom perceivers adhere topreventive regimens better than poor perceivers do. To bet-ter understand symptom perception, we also explored pa-tient characteristics that might contribute to poor symptomperception. We were interested in the significance of illnessidentity factors, including perceived illness severity, control-ling for FEV1 (a measure of objective illness severity), yearswith an asthma diagnosis, sex, and socioeconomic status(income). Studies have shown an association of sex withsymptom perception (van Wijk & Kolk, 1997), but the lit-erature on socioeconomic status in perceptual accuracy hasbeen limited (Yoos & McMullen, 1999). Nonetheless, thesevariables were included for analyis.

Methods

This was a cross-sectional descriptive study with dailymeasures over 4 weeks, designed to evaluate relationshipsamong symptom perception, illness identity, and adherencerates.

Study Participants and SettingA convenience sample of adults with mild to se-

vere asthma, ages 25 to 60, was recruited from threeasthma/allergy specialty clinics. A power analysis for a de-sired power of .8 and α=.05, based on six independent vari-ables and a medium effect, showed 97 participants neededfor multiple regression analysis (Green, 1991). Allowing for20% attrition, 120 persons initially were recruited. Six en-rolled but subsequently did not meet eligibility criteria; 17did not complete the study.

Inclusion criteria included a physician’s diagnosis ofmild, moderate, or severe persistent asthma; a prescrip-tion for Advair� (a dry, powder-based inhaled corticosteroid(ICS) medication dispensed from a device that maintains

Journal of Nursing Scholarship Third Quarter 2006 293

Symptom Perception

count of remaining doses); and ability to read and followinstructions in English. Exclusion criteria included comor-bidities of heart failure or chronic obstructive pulmonarydisease and current smoking.

MeasuresSymptom perception was defined operationally as a per-

ceptual accuracy score (PAS). A PAS was calculated for eachparticipant by correlating their peak expiratory flow rate(PEFR) scores with their subjective breathlessness (dyspnea)scores, measured with the single-item Intensity Numeric Rat-ing Scale (INRS). The PEFR was obtained by using a “Per-sonal Best Peak Flow Meter,” which has a range of 60 to 810liters/minute (L/min). The manufacturer reports accuracyat ±10% or 20 L/min, reproducibility of ≤5%, or 10 L/minand inter-device variability of ≤10% or 20 L/min (Respiron-ics, 2001). The INRS ranges from 0 (not noticeable) to 10(extremely severe) and was selected because it simplifiesscoring and avoids potentially inflated results related to in-creased variability often found with a VAS (Aaronson et al.,1999). Youngblut and Casper (1993) provided a compellingargument for the validity of single-item measures.

Because participants rated their subjective perceptionsof dyspnea intensity with the INRS twice daily for 4 weeks,and did additional assessments whenever they perceivedsymptoms of dyspnea, a minimum of 28 PEFR and INRSscore pairs was established for calculating the PAS correla-tion. Pearson correlations of −.30 or weaker represent lowaccuracy (poor perception), −.31 to −.59 represent moder-ate accuracy, and −.60 or stronger represent high accuracy(both considered good perception). This is consistent withthe values of PAS scores determined by Yoos and McMullen(1999) in their study of symptom perception in childhoodasthma.

Illness Identity was assessed by two measures: a single-item indicator of perceived asthma severity and a theoret-ically based measure, the Revised Illness Perception Ques-tionnaire (IPQ-R; Weinman, Petrie, Moss-Morris, & Horne,1996; Moss-Morris et al., 2002). For perceived asthmaseverity, participants identified their perception of the sever-ity of their asthma on a scale of 1 (very mild) to 10 (verysevere). Test-retest reliability of this measure in this studywas strong (r=.70) as measured pre- and poststudy 4weeksapart. Construct validity of the item was supported by apositive correlation with the consequences subscale of theIPQ-R (r=.48) and a negative correlation between the FEV1

and perception of severity (r=−.34).The IPQ-R is a revision of the Illness Perception Ques-

tionnaire (IPQ; Lau et al., 1989; Leventhal & Nerenz, 1985).It includes seven subscales: timeline: acute/chronic, (likelyduration of asthma); timeline: cyclical, (perception of illnessvariability); consequences (effect of asthma on person’s life);personal control (self-efficacy and personal control over thedisease); treatment control (belief in efficacy of the treat-ment or provider recommendations); emotional represen-tations (affective response to asthma); and illness coherence(patient’s overall comprehension of asthma). Participants re-

Table 1. Means, Standard Deviations, and ReliabilityCoefficients of Scales

Scale Mean SD N a α test-retest r

Perceived Asthma Severity 5.12 2.03 114 n/a .70 (n=91)b

IPQ-R subscalesTimelinec 4.00 .83 113 .86 .54 (n=90)Consequencesc 3.17 .79 114 .78 .76 (n=91)Personal controlc 4.16 .53 114 .79 .61 (n=91)Treatment controlc 4.07 .51 114 .65 .62 (n=91)Illness coherence 3.67 .87 114 .91 .75 (n=91)Time cyclicc 3.13 .76 112 .73 .48 (n=89)Emotional representation 2.22 .88 112 .91 .79 (n=89)

MARS 4.25 .60 114 .83 .74 (n=91)

Note. aCronbach alphas were determined from prestudy data of 114 eligible partic-ipants. bTest-retest reliability scores required poststudy data (97 participants com-pleted the study); missing data resulted in n’s ranging from 89 to 91. cSubscalesincluded in this analysis.

sponded on a 5-point Likert scale to 38 statements aboutasthma. Scores were reverse coded as appropriate, and sub-scale mean scores were calculated. Test-retest reliability andCronbach alphas for each subscale are shown in Table 1.For this study we used the timeline-cyclical, consequencesand control subscales.

Adherence was measured in two ways, inhaler use (dosecounts) and by self-report. The Advair� “diskhaler” recordseach time it is used, and adherence was calculated by divid-ing the number of counted doses by the number of prescribeddoses and multiplying by 100 (percentage Advair�). Goodor adequate adherence was defined as using 80% or moreof the prescribed doses (Rapoff, 1999).

The Medication Adherence Report Scale (MARS) wasused as the self-report measure and is a nine-item Likert-typescale to assess nonadherent behavior in a nonthreateningmanner (Horne & Weinman, 2002). Response options rangefrom 1 (very often) to 5 (never). Examples of items are: “I al-ter the dose” and “I forget to use it.” Items are reverse-scoredas appropriate and a mean score is determined. The MARShas shown adequate internal consistency with a Cronbach’salpha of .85 (Horne & Weinman). In this study, the Cron-bach’s alpha was .83, and the pre-post study MARS werehighly correlated (r=.74), indicating good stability. The per-centage Advair� count also was moderately correlated withthe MARS post-study score (r=.53), indicating construct va-lidity of the MARS.

ProceduresApproval for the study was granted by an institu-

tional review board before data collection. After potentialparticipants were informed of the purpose of the study andsigned an informed consent, they each were provided apeak flow meter and instructed on proper technique for itsuse. Participants performed the maneuver until they demon-strated acceptable technique. Participants were instructed torate their subjective perception of breathlessness intensity


Symptom Perception

twice daily (morning and evening), before measuring theirpeak flow rate and using their inhalers. Additional assess-ments were done whenever they perceived symptoms of dys-pnea.

Four weekly booklets for daily data recording, withstamped, addressed envelopes, were given to each partici-pant. They were instructed how to record the data daily andto return a completed data booklet at the end of each week.If booklets were not returned in a timely manner, reminderphone calls were made.

The MARS, the IPQ-R, the subjective severity of asthmasingle-item indicator, and the demographic form were com-pleted at the initial interview. All except the demographicform were completed again at the end of the 4-week data-collection period. Final Advair� counts also were determinedat the end of the 4-week period.

Findings

Sample CharacteristicsSeventy-eight percent of the participants were women;

the average age was 44.8 years (SD=9.27). The sample waspredominantly Caucasian (83.5%), with 10.3% Hispanic,3.1% African American, and 3.1% other (Native Americanand Asian). The majority (87%) had household incomesgreater than $30,000/year. They also were well educated:48% had college degrees or higher and only two did nothave a high school diploma. Age, height, weight, perceivedasthma severity, years with asthma, spirometry results (e.g.,FEV1%), PEFR, and morbidity indicators were comparedacross the three sites. The only significant difference betweensites was the FEV1% (Table 2). The FEV1%, rather thanthe FEV1, was used because it yields a standardized value tocompare across people.

Of the 97 participants who completed the study, 94 per-ceptual accuracy scores (PAS) were calculated. The meanPAS was −.45. Sixty-two participants (67%) had a PASstronger than −.30, and were considered “good perceivers;”32 participants (34%) were “poor” perceivers, with a PASof −.30 or weaker. Three had no variability in their INRSscores and maintained PEFR at a level where symptomswould not normally be perceived, so a PAS could not be de-termined. Two others had no variability in their INRS scores(consistently rating zero symptoms), but had PEFR that var-ied and were well into the range where symptoms shouldbe perceived. For these two participants, a PAS of −.20 wasassigned, indicating poor perception of symptoms.

Eighty-two percent of the participants took 80% ormore of their prescribed doses. The mean percentage foradherence was 90.2% with a SD=12.8.

Analyses were done to test the hypothesis that accu-rate, or good, symptom perceivers will adhere to preventiveregimens better than poor perceivers do. An independentt test comparing adherence of good and poor perceivers,measured by the PAS, showed no significant difference inpercentage of Advair� taken: t(86.93)=−1.85, p<.07 (equal

Table 2. Comparison of the Sample From Three Clinic Sites

Site 1 Site 2 Site 3n=61-65a n=15-18a n=12-14a

Mean (± SD) Mean (± SD) Mean (± SD) F (df ) p

Age 45.2 (9.2) 44.5 (10.1) 43.0 (9.1) .31(2,93) .734

Height 65.7 (3.9) 64.4 (3.4) 66.4 (4.9) 1.02(2.94) .366

Weight 188.8 (44.0) 177.7 (33.1) 200.9 (51.7) 1.13(2,94) .327

Asthmaseverity 4.8 (2.2) 5.6 (2.0) 5.2 (1.8) .94(2,94) .393

Years withasthma 14.9 (16.2) 16.8 (13.2) 16.6 (13.4) .16(2,94) .854

FVC% 86.6 (14.9) 88.5 (11.7) 88.4 (17.9) .17(2,94) .843

FEV1% 92.4 (19.3) 80.1 (14.6) 83.6 (19.6) 3.77(2,94) .027

FEV25–75% 84.7 (35.3) 62.8 (25.7) 75.6 (37.2) 3.01(2,94) .054

FEV1/FVC 80.9 (9.4) 78.3 (12.1) 78.8 (8.8) .66(2,93) .520

PEFR 403 (106) 432 (91) 439 (113) .96(2,90) .385

Hospitalvisits .05 (.28) .00 (0) .07 (.27) .37(2,94) .695

ER visits .23 (1.17) .06 (.24) .07 (.27) .32(2,94) .727

Office visitsb .62 (1.17) .33 (.69) .86 (1.83) .75(2,92) .475

Missed work 2.5 (11.4) .47 (1.3) .00 (0) .55(2,90) .578

Note. aData were missing for some variables. bViolated Levine test for homo-geneity of variance at p<.05.

variances not assumed). Similarly, no significant difference(t(92)=−1.20, p=.24) was found between the groups on theprestudy MARS self-rating; nor was a significant differencefound between good and poor perceivers on adherence asmeasured by the post-study MARS: t(86)=−1.75, p=.08. Be-cause of near-significant counter-intuitive results, i.e., thatgood perceivers might have poorer adherence, analyses ofPEFR variability were conducted.

Originally, we only considered the PEFR in calculatingthe PAS, not the issue of relative variability in PEFR withineach participant. However, given the data and our goal tobetter understand symptom perception in relation to asthmatreatment management, it became a logical next step. If peo-ple with poorer adherence to controller medications experi-ence wider fluctuations of respiratory status, they might ex-perience more symptoms and, consequently, have a strongerperceptual accuracy score.

To determine if PEFR variability was a factor in symp-tom perception, a percentage PEFR value was computedfor each particpant. This percentage of peak expiratoryflow rates (%PEFR) was calculated by dividing the rangeof PEFR (maximum minus minimum for each person), bythe person’s maximum value (personal best). The%PEFRindicates the amount of variability in respiratory function aparticipant experienced during the 4 weeks of data collec-tion. The mean%PEFR was 25.56. Because symptoms aregenerally experienced when the PEFR varies at least 20%from a personal best (such as PEFR value of 400 with a “per-sonal best” of 500), 20% was selected as the%PEFR cut-off


Symptom Perception

point. This approach resulted in 42 participants with PEFRvariability less than or equal to 20% and 55 participantswith variability greater than 20%.

Independent t tests were conducted to determine if PEFRvariability was significantly related to the PAS and to themeasures of adherence. Participants with higher PEFR vari-ability had stronger symptom perception scores (PAS) com-pared to those with PEFR variability ≤20%: t(92)=3.17,p=.002. Those with PEFR variability ≤20% also had higheradherence as measured by percentage Advair�: t(85.38)=2.09,p=.04; equal variances not assumed. The relationship be-tween PEFR variability and self-reported adherence at post-study was not statistically significant.

Illness Identity and Symptom PerceptionAnalyses were done to test the research question: Do

illness identity factors, including perceived illness severity,affect symptom perception, controlling for FEV1%, yearswith an asthma diagnosis, sex, and socioeconomic status(income). The four IPQ-R subscales (personal and treatmentcontrol, consequences, and timeline-cyclic), the single itemfor perceived illness severity, and the other independent vari-ables were assessed for potential collinearity. Because thepersonal and treatment control subscales were highly corre-lated (r=.57, p<.001), they were combined. The perceivedasthma severity scale was correlated with the consequencessubscale (r=.50, p<.001) and was dropped from these anal-yses because we also had FEV1% as an objective measure ofasthma severity. Finally, in light of the above analyses, weadded PEFR variability to these analyses.

A multiple regression analysis was conducted with thefollowing variables: personal/treatment control subscale,consequence subscale, timeline-cyclic subscale, PEFR vari-ability (%PEFR), FEV1%, years with asthma, sex, and in-come. The linear combination of these variables was sig-nificantly related to symptom perceptual accuracy: R2=.34,adjusted R2=.27, F(8,77)=5.01, p<.001. The PEFR variabil-ity was a particularly strong explanatory variable of per-ceptual accuracy: β=−.374, t=−3.932, p<.001. The IPQ-R subscales (personal/treatment control, consequence, andtimeline-cyclic), indicating the participant’s illness identity,also were significant explanatory variables of perceptual ac-curacy (see Table 3).

Discussion

The hypothesis that accurate, or good, symptom per-ceivers will adhere to preventive regimens better than dopoor perceivers was not supported by the results of thisstudy. In part, this result may have been because of restrictedvariance in the measures of adherence. However, becausethe trends in our findings were in the opposite directionfrom what we hypothesized, we introduced the concept ofPEFR variability and found that it was an important con-cept for understanding perceptual accuracy, likely becausesymptoms are easier to detect when respiratory function is

Table 3. Summary of Multiple Regression Analysis forVariables Explaining Perceptual Accuracy Scores

Variable B SE B β t p

Constant .815 .337 2.423 .018Years with asthma −.004 .002 −.197 −2.046 .044Income −.021 .017 −.115 −1.206 .231Sex −.005 .065 −.009 −.085 .933FEV1 percent .001 .001 .072 .731 .467Personal/treatment control −.130 .059 −.225 −2.228 .029Consequence −.072 .033 −.213 −2.180 .032Timeline cyclical −.069 .033 −.203 −2.118 .037Peak flow variability −.783 .199 −.374 −3.932 .000

(%PEFR)

Note. F (8,77)=5.014, p<.001; R2=.34; Adjusted R2=.27; SE of Estimate=.22.

more varied. Not only was PEFR variability positively re-lated to perceptual accuracy, but it was better than percep-tual accuracy for discriminating good adherers from pooradherers.

In a pilot study of 18 adolescents, Wamboldt (1998)found those with higher perceptual accuracy scores were lessadherent with oral corticosteroid therapy. Wamboldt sug-gested that these findings might indicate that good perceiversare less likely to take their controller medications whenasymptomatic. The present study indicates an alternate ex-planation: Patients who do not consistently take their con-troller medication are more likely to experience increasedvariability in pulmonary function and thus are more likely toaccurately perceive their symptoms. Although we found nodifferences in adherence between good and poor perceiverswhen measured with the PAS, those with low peak expira-tory flow rate variability (≤20%) had significantly higheradherence rates as indicated by the percentage of Advair�

over a 4-week period.These findings illustrate difficulties with cross-sectional

designs. Wamboldt (1998) interpreted her findings that goodsymptom perception led to more frequent dose-skippingwhen asymptomatic. Our analyses could be interpreted thatgreater symptom variability leads to better symptom percep-tion. Possibly, poorer adherence is what leads to the greatervariability in the first place. Determining causation is notpossible without specification of a true starting point.

Still, accurate symptom perception is important forgood self-management of asthma. The variables identifiedin this study as statistically and conceptually significantin explaining variance in symptom perception are consis-tent with the symptom interpretation model (SIM). The in-put of the physiological change associated with increasedvariation of peak expiratory flow had the strongest ef-fect on symptom perception. This makes conceptual sense.Physiologically, large changes in airway flow, indicted bygreater variability of peak expiratory flow, are more likelyto reach a perceptual threshold.


Symptom Perception

The SIM process of interpretation was measured in thisstudy with the illness identity variables. All three subscalesof the IPQ-R were significant contributors to the explainedvariance in symptom perception, indicating that those whoassociated more severe consequences with asthma, who per-ceived more variability in the course of their asthma, andwho reported a stronger belief in their ability to manageasthma symptoms, were more likely to have better asthmasymptom perceptual accuracy. In addition, more experiencewith asthma (i.e., years with asthma) also was associatedwith better perceptual accuracy, possibly because of a well-established illness identity over time.

Finally, the generally high adherence to controller medi-cations in this sample may be based on several factors. First,all participants were clients of asthma and allergy special-ists. Although this consistency in provider specialization de-creased variability, it also may have contributed to less vari-ability in how well asthma was controlled in this sample. Be-cause they were receiving asthma specialty care, they weremore likely to believe their diagnosis and, thus, may havebeen more likely to adhere to their medications. Second, oursample was predominantly Caucasian, well-educated, and ofreasonably substantial financial means. These demographicvariables consistently have been associated with more pos-itive health behaviors. Third, the repeated measurement ofperceived symptoms and PEFR, necessary for data collec-tion in this study, unavoidably also was interventional. Theact of focusing on and recording their symptoms and PEFRmay have improved their adherence and contributed to agenerally high adherence rate.

Additional research on the complex relationshipsamong symptom variability, symptom perception accuracy,and adherence to asthma controller medications is clearly in-dicated. In particular, studies with more ethnically and eco-nomically diverse samples, as well as study participants whoreceive their asthma care from a variety of providers, areneeded. Prospective study is also important, perhaps withpeople newly diagnosed with asthma, which could lead toknowledge about the relationship between symptom percep-tual accuracy and treatment adherence.

Conclusions

Although findings from this study did not support theoriginal hypothesis that perceptual accuracy would be posi-tively associated with treatment adherence, our findings didshow the importance of respiratory function variability (asmeasured by peak flow variability) for symptom perceptionaccuracy and raised questions about the direction of cau-sation between symptom perception accuracy and asthmamedication adherence. The importance of illness identity,indicated by perceptions of the consequences of asthma, thecyclic nature of the illness, and beliefs about personal andtreatment control for asthma, as well as respiratory func-tion variability and the length of time with an asthma di-agnosis, all contributed significantly to symptom perceptualaccuracy.

References

Aaronson, L.S., Teel, C.S., Cassmeyer, V., Neuberger, G.B., Pallikkathayil,L., Pierce, J., et al. (1999). Defining and measuring fatigue. Image: Jour-nal of Nursing Scholarship, 31(1), 45–50.

American Lung Association. (2005). Trends in asthma mor-bidity and mortality. Retrieved December 1, 2005, fromhttp://www.lungusa.org/site/pp.asp?c=dvLUK9O0E&b=22884

Banzett, R.B., Dempsey, J.A., O’Donnell, D.E., & Wamboldt, M.Z. (2000).Symptom perception and respiratory sensation in asthma. AmericanJournal of Respiratory Critical Care Medicine, 162(Pt. 1), 1178–1182.

Centers for Disease Control. (1996). Asthma mortality and hospitalizationamong children and young adults—United States, 1980–1993. Morbid-ity and Mortality Weekly Report (MMWR), 45, 350–353.

Green, S.B. (1991). How many subjects does it take to do a regressionanalysis? Multivariate Behavioral Research, 26(3), 499–510.

Harver, A., & Mahler, D.M. (1998). Perception of increased resistance tobreathing. In H. Kostes & A. Harver (Eds.), Self management of asthma(pp. 147–193). New York: Marcel Dekker.

Horne, R., & Weinman, J. (2002). Self-regulation and self-management inasthma: Exploring the role of illness perceptions and treatment beliefsin explaining non-adherence to preventer medication. Psychology andHealth, 17(1), 17–32.

Jain, P., & Kavuru, M. (2000). Peak expiratory flow vs. spirometry in apatient with asthma. Respiratory Care, 45(8), 969–970.

Julius, S.M., Davenport, K.L., & Davenport, P.W. (2002). Perception ofintrinsic and extrinsic respiratory loads in children with life-threateningasthma. Pediatric Pulmonology, 34(6), 425–433.

Lau, R., Bernard, J. M., & Hartman, K. A. (1989). Further explorations ofcommon sense representations of common illness. Health Psychology, 8,195–219.

Lehrer, P., Feldman, J., Giardino, N., Song, H.S., & Schmaling, K. (2002).Psychological aspects of asthma. Journal of Consulting and Clinical Psy-chology, 70(3), 691–711.

Leventhal, H., & Nerenz, F. (1985). The assessment of illness cognition.In P. Karoly (Ed.), Measurement strategies in health psychology (pp.517–555). New York: Wiley.

Magadle, R., Berar-Yanay, N., & Weiner, P. (2002). The risk of hospital-ization and near-fatal and fatal asthma in relation to the perception ofdyspnea. Chest, 121(2), 329–333.

Moss-Morris, R., Weinman, J., Petrie, K. J., Horne, R., Cameron, L. D., &Buick, D. (2002). The Revised Illness Perception Questionnaire (IPQ-R).Psychology and Health, 17(1), 1–16.

National Asthma Education and Prevention Program. (1997). NAEPP ex-pert panel report 2: Practical guide for the diagnosis and managementof asthma. Bethesda, MD: Author.

National Asthma Education and Prevention Program. (2002). NAEPPexpert panel report guidelines for the diagnosis and management ofasthma—Update on selected topics 2002. (NIH Publication No. 02–5075). Bethesda, MD: National Institute of Health.

National Center for Health Statistics. (2003). Asthma prevalence, healthcare use and mortality, 2000–2002. Retrieved March 2, 2003, from http://www.cdc.gov/nchs/products/pubs/pubd/hestats/asthma/asthma.htm

Rapoff, M.A. (1999). Adherence to pediatric medical regimens. New York:Kluwer Academic/Plenum.

Respironics. (2001). Personal Best Peak Flow Meter: Instructions for useCedar Grove, NJ: Author.

Teel, C.S., Meek, P., McNamara, A.M., & Watson, L. (1997). Perspectivesunifying symptom interpretation. Image: Journal of Nursing Scholarship,29(2), 175–181.

van Wijk, C.M., & Kolk, A.M. (1997). Sex differences in physicalsymptoms: The contribution of symptom perception theory. Social Sci-ence & Medicine, 45(2), 231–246.

Wamboldt, M.Z. (1998). A wheeze by any other name is (not) the same:The role of symptom perception in asthma. Journal of Asthma, 35(2),133–135.

Weinman, J., Petrie, K. J., Moss-Morris, R., & Horne, R. (1996). The ill-ness perception questionnaire: A new method for assessing the cognitiverepresentation of illness. Pyschology and Health, 11, 431–445.

Yoos, H.L., & McMullen, A. (1999). Symptom perception and evaluationin childhood asthma. Nursing Research, 48(1), 2–8.

Youngblut, J.M., & Casper, G.R. (1993). Single-item indicators in nursingresearch. Research in Nursing & Health, 16, 459–465.


symptom perception and adherence to asthma controller medications

Documents