Review article
`Supported by a grant from Zeneca Pharmaceuticals
`
`Nonadherence in asthmatic patients: is there a
`solution to the problem?
`
`Bruce Bender, PhD*†; Henry Milgrom, MD*‡; and Cynthia Rand, PhD§
`
`Learning Objectives: Reading this article will reinforce the reader’s awareness
`of the relationship between adherence and treatment outcome, of the causes of
`nonadherence, of methods of measurement, and of steps toward successful inter-
`vention.
`Data Sources: Articles on adherence to asthma therapy were reviewed. A
`MEDLINE database using subject keywords was searched from 1990 through 1997.
`Study Selection: Pertinent articles were chosen, with preferential presentation of
`results from controlled studies.
`Results: There is no evidence of recent improvement in the rates of nonadher-
`ence, and patients continue on average to take about 50% of prescribed medication.
`Nonadherence assessment is most accurate when it can be measured objectively, and
`relies neither on patient report nor physician estimate. The consequences of non-
`adherence are measured in patient suffering, financial cost, and serious compromise
`of clinical trial outcomes. Underlying causes of nonadherence are traced to char-
`acteristics of the disease, treatment, patient, and caregiver system.
`Conclusion: Improved adherence will lead to improved disease control, but only
`if medical care systems encourage and support the allocation of sufficient resources
`to allow barriers to self-management to be discussed and solutions negotiated.
`Attempts to improve adherence outside of the caregiver-patient relationship are less
`likely to succeed. Special programs for difficult-to-manage patients are necessary to
`change behavior, although significant illness improvement and cost savings are
`likely to result.
`
`Ann Allergy Asthma Immunol 1997;79:177–86.
`
`INTRODUCTION
`That medication nonadherence is a
`common problem interfering with ef-
`fective asthma management is well-
`established. Still, the extent, causes,
`and implications of nonadherence are
`not universally appreciated, and at-
`tempts to intervene have been at best
`
`From the Departments of * Pediatrics, † Psy-
`chiatry, and ‡ Medicine, National Jewish Medi-
`cal and Research Center and the University of
`Colorado Health Sciences Center, Denver, Col-
`orado; and the § Division of Pulmonary and
`Critical Care Medicine, Johns Hopkins School
`of Medicine, Baltimore, Maryland.
`Received for publication June 10, 1997.
`Accepted for publication in revised form July
`25, 1997.
`
`only partially successful. This manu-
`script will discuss the relationship be-
`tween adherence and treatment out-
`come and explore contributing factors
`and underlying causal mechanisms. Fi-
`nally, methods
`for measuring and
`changing patient adherence behaviors
`will be discussed.
`Adherence is the extent to which a
`patient follows a reasonable treatment
`plan that has been prescribed for them
`by a qualified caregiver. While intelli-
`gent nonadherence has been noted in
`cases where a treatment was detrimen-
`tal to a patient who recognized this fact
`when his caregiver did not, the term
`adherence is adopted here with the un-
`derstanding that most nonadherence
`
`undermines a patient’s health and well-
`being. To argue otherwise is to ignore
`overwhelming evidence. In a review of
`ten pediatric asthma adherence studies,
`medication adherence averaged 48%.1
`Regardless of whether adherence is
`measured as serum theophylline levels
`at clinic visits,2 percent of prescribed
`medication taken,3 days of medication
`adherence,4 or percent of patients who
`fail to reach a clinically estimated ad-
`herence minimum,5 rates of nonadher-
`ence among asthmatic patients typi-
`cally range from 30% to 70%.
`Adherence must not be exclusively
`defined in reference to medication use.
`Patients who attempt to use their in-
`haled medication but do not adhere to
`good inhalation technique experience
`much less medication benefit.6 Current
`Guidelines for the Diagnosis and Man-
`agement of Asthma7 conclude that con-
`trol of severe asthma also depends
`upon peak flow meter use to monitor
`lung function, avoidance of allergens
`and irritants, and appropriate commu-
`nication with the physician’s office.
`One study disclosed that patients failed
`to use their meter on 37% of days,
`while reporting such omission on their
`diary cards that nonuse occurred on
`only 7% of study days.8 Other behav-
`iors that contribute to good health are
`important, although difficult to define
`and measure.
`
`IMPACT OF NONADHERENCE
`While a variety of solutions to patient
`nonadherence have been offered, their
`impact has been small and their em-
`ployment limited to very few treatment
`settings. At present, evidence indicates
`
`VOLUME 79, SEPTEMBER, 1997
`
`177
`
`Liquidia's Exhibit 1032
`Page 1
`
`

`

`that, overall, the problem of nonadher-
`ence to asthma treatment is not im-
`proving. The consequences of nonad-
`herence continue to be measured in
`personal, financial, and research costs.
`Personal Cost
`While decreased adherence in some
`patients may not compromise disease
`control, the consequence of nonadher-
`ence for many patients is illness exac-
`erbation. Children who did not adhere
`to therapy demonstrated significantly
`more wheezing, greater variability in
`peak flow rates, and lower asthma con-
`trol scores,9 while adults whose airway
`obstruction failed to resolve were
`markedly less compliant
`than those
`who improved.10 More tragically, non-
`adherence has been associated with
`asthma-related deaths in children, par-
`ticularly where psychologic dysfunc-
`tion has been observed in the patient or
`the patient’s family.11
`Growing evidence reveals that many
`severely asthmatic patients are dramat-
`ically nonadherent. Across various
`chronic diseases, adherence improves
`as disease severity increases from mild
`to moderate, but appears to reverse
`with severe illness.12 Hospitalizations
`and emergency room visits, it might be
`assumed, would dramatically boost ad-
`herence motivation because they are
`frequently unpleasant and costly expe-
`riences that signal to patient and care-
`giver that a serious health decline has
`occurred. Strikingly, this assumption is
`frequently incorrect. Nonadherence is
`often high among patients who appear
`in the emergency room or hospital, and
`there is no evidence that such contacts
`with urgent care serve to increase and
`sustain subsequent adherence. Theo-
`phylline-treated asthmatic patients pre-
`senting at an emergency room have
`been found to have subtherapeutic
`theophylline levels.13 In a 90-day lon-
`gitudinal study, eight children requir-
`ing urgent care visits resulting in oral
`steroid bursts (two of whom required
`subsequent hospitalization) had been
`dramatically less adherent with a daily
`regimen of inhaled steroids than a
`
`group of 16 patients with stable symp-
`tom control.3
`Perhaps the most revealing evidence
`regarding the consequences of nonad-
`herence in severely asthmatic patients
`emerges from studies attempting to al-
`ter these patterns. A number of studies
`provided a psychoeducational
`inter-
`vention without altering availability or
`cost of medical care and medications.
`The degree to which the course of se-
`vere asthma was altered in these pro-
`grams reflects the extent
`to which
`inadequate healthcare behavior con-
`tributes to poorly controlled asthma.
`Adult patients participating in a
`7-week asthma self-management class
`were found to have significantly re-
`duced asthma symptoms at 1-year fol-
`low-up.14 In another study, 104 adults
`with a history of asthma-related hospi-
`talizations participated in a program
`emphasizing teaching patients self-
`management
`strategies
`in case of
`marked asthma exacerbation, resulting
`in a two-thirds reduction in readmis-
`sions.15
`Financial Costs
`Asthma is a costly disease. The direct
`and indirect cost of asthma in 1990
`was estimated to be $6.2 billion. The
`$3.6 billion direct medical costs in-
`cluded physician visits and medica-
`tion, but the largest portion (almost $3
`billion) resulted from emergency room
`visits and hospitalizations.16 Since only
`a small proportion of asthmatic pa-
`tients require hospitalization, much of
`the cost of asthma is created by a rel-
`atively small group of patients. Ap-
`proximately 5% of asthmatic patients
`account for more than 70% of the total
`cost of asthma.17 Many of those asth-
`matic patients who present at the emer-
`gency room or hospital have been there
`before.16 That inadequate health care
`behavior results in increased cost is
`evidenced by the finding that a few
`programs targeting self-management
`behavior achieved significant subse-
`quent savings through reduced hospi-
`talizations
`and
`emergency
`room
`use.14,18,19
`
`Clinical Investigation Costs
`Nonadherence in clinical trials can di-
`lute treatment effects and result in er-
`roneous conclusions. Just as in clinical
`practice, study patients are frequently
`nonadherent with their treatment.3,20
`Clinical trial nonadherence can encom-
`pass a variety of behavioral absences,
`including failure to take study medica-
`tion, failure to perform other protocol-
`dictated tasks such as completion of
`diary cards, and failure to attend study
`visits. Dropping out of a study is a
`relatively extreme act of nonadherence
`often preceded by medication nonad-
`herence and sometimes resulting in
`distortion of outcome data.21 Study de-
`parture is an overt and obvious behav-
`ior that can be addressed in statistical
`analyses of outcome data.
`Other forms of nonadherence, such
`as recording of fabricated data on diary
`cards, are much more difficult to rec-
`ognize and therefore have potential for
`corrupting the study. One longitudinal
`study of the treatment of chronic ob-
`structive lung disease disclosed that in
`an apparent attempt to conceal nonad-
`herence 14% of participants were
`“dumping” large amounts of aerosol-
`ized medication shortly before sched-
`uled visits while reporting adherence
`on their diary cards.20
`If undetected nonadherence signifi-
`cantly reduces drug use in a trial, the
`reported effectiveness of the trial med-
`ication will be seriously diluted and
`may result in unnecessarily high rec-
`ommended doses. Ironically, adherent
`patients may consequently have in-
`creased exposure to potential side ef-
`fects. If adherence rates differ for two
`equally effective drugs, the investiga-
`tors may erroneously ascribe to phar-
`maceutical superiority a clinical differ-
`ence partially or completely due to
`behavioral differences between study
`groups.22 Without sufficient monitor-
`ing of adherence, clinical trial results
`continue to be based upon an inappro-
`priate average of adherence and fail to
`take advantage of the opportunity to
`increase insight into dose-response re-
`lationships introduced by varying lev-
`els of adherence.
`
`178
`
`ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY
`
`Liquidia's Exhibit 1032
`Page 2
`
`

`

`CORRELATES OF
`NONADHERENCE
`The causes of nonadherence are many
`but generally fall into one of three cat-
`egories.
`Disease and Treatment-Related
`Factors
`Patients are less likely to adhere to
`their treatment regimen if their disease
`is either mild or severe,12 a troubling
`finding particularly in the case of pa-
`tients with poorly controlled asthma.
`Adherence is further undermined in the
`presence of chronic illness requiring
`prolonged treatment, where the pre-
`scribed medications are used prophy-
`lactically, and where the consequences
`of cessation of treatment are delayed.23
`Medication expense and side effects
`often deter patients.12 When medica-
`tions are difficult to take, adherence
`declines; one study of inhaled medica-
`tion use in asthmatic children disclosed
`71% adherence with twice-daily dos-
`ing, decreasing to 34% with three-
`times-a-day and 18% with four-times-
`a-day dosing24; however, other studies
`have not replicated this dose-response
`pattern of adherence.4 Still, simplifica-
`tion of the treatment regimen tends to
`foster improved adherence. Once daily
`dosing with theophylline tablets re-
`sulted in dramatically higher adher-
`ence rates than twice daily use of in-
`haled corticosteroids or
`cromolyn
`sodium in asthmatic adolescents and
`adults.25 Some patients simply do not
`like taking inhaled medications, and
`many fail to master the necessary skill
`required for effective actuation-to-in-
`halation coordination required with
`these drugs.26 Unfortunately, most of
`these
`factors,
`including long-term
`treatment, delayed cessation of treat-
`ment consequence, expense, side ef-
`fects, skill requirements, and frequent
`dosing schedules, all characterize the
`treatment of significant asthma.
`Patient-Related Factors
`Many patients are nonadherent, and no
`simple profile characterizes all or even
`most nonadherent patients. Failure to
`adhere to treatment regimen can be
`found in patients of varying demo-
`
`graphic, psychologic, or illness pro-
`files. Some patient characteristics,
`however, are correlated with medica-
`tion adherence,
`indicating the exis-
`tence of specific subgroups of nonad-
`herent
`patients. Not
`surprisingly,
`reduced intelligence has been associ-
`ated with poor adherence. Increased
`age, on the other hand, has not; while
`elderly individuals might be expected
`to have difficulty tracking and remem-
`bering their medication regimen, most
`studies have failed to disclose reduced
`adherence in this population relative to
`younger adults.27,28
`Psychopathology is the clearest pa-
`tient characteristic associated with
`medication
`nonadherence.28–30 De-
`creased capacity to maintain a consis-
`tent regimen of disease self-manage-
`ment
`is
`evident
`in
`not
`only
`psychiatrically disturbed individuals,
`but also dysfunctional families.2,31 Dis-
`tressed or psychologically unstable
`parents may not provide the structure
`and support necessary to ensure adher-
`ence of their children. In many such
`cases, children are given inappropriate
`responsibility for their own medical
`care; when parents are not committed
`to following a prescribed asthma man-
`agement program, it is unrealistic to
`expect good adherence from the child.
`Serious psychologic disorder can re-
`sult in denial and create a particularly
`dangerous situation. A pattern of poor
`self-management and significant psy-
`chologic dysfunction in patients and
`their families has been identified in
`children and adolescents who died of
`asthma.11 A history of erratic asthma
`management, conflict between parent
`and child, frequent changes of health
`care provider, and child depression or
`anger are signals that standard medical
`care alone cannot adequately address
`nonadherence. Depression and loss of
`faith in the value of therapy are com-
`mon in patients who willfully fail to
`comply and manipulate their asthma
`for secondary gain.32,33
`Even in the absence of psychologi-
`cal dysfunction, adherence is largely
`mediated by the psychologic outlook
`of the patient or the patient’s parents.
`The Health Belief Model34,35 was for-
`
`mulated to explain how a patient’s ex-
`periences, perceptions, and beliefs
`guide their understanding of, and re-
`sponse to, their disease. The Health
`Belief Model holds that patients fre-
`quently conduct their own cost-benefit
`analysis with regard to treatments pro-
`posed by their health care provider.
`Patients are more likely to adhere to a
`prescribed treatment if they perceive
`their illness as significant, and if they
`believe that
`the proposed treatment
`will be effective without adverse con-
`sequence such as medication side ef-
`fects, financial sacrifice, or lifestyle
`change. These beliefs are greatly af-
`fected by the patient’s experiences and
`information obtained from acquaintan-
`ces and the public media. Awareness
`of controversies surrounding first-line
`asthma
`therapies—corticosteroids,
`theophylline, and, most recently, beta-
`agonists—contribute to reluctance and
`increase nonadherence. Failure of one
`therapy to effectively manage symp-
`toms may create skepticism toward
`successive therapies. Patients com-
`monly consult more than one caregiver
`for their asthma;
`lack of agreement
`between caregivers further undermines
`the belief that the effort, cost, and in-
`convenience of asthma treatment are
`justified. Once established, health be-
`liefs are not easily altered.28 If a pa-
`tient’s asthma is misdiagnosed, initial
`treatments are ineffective, and educa-
`tion about the disease is incomplete or
`inaccurate, the consequent health be-
`liefs formulated by the patient will cre-
`ate a barrier to subsequent effective
`disease management.
`Conversely, a positive orientation
`toward health not only strengthens
`medication adherence but also leads to
`other important if difficult-to-measure
`health behaviors. Evidence for the con-
`tribution of positive health care behav-
`ior, aside from medication usage, is
`seen in a study of almost 4000 patients
`who had experienced myocardial in-
`farction.36 The 5-year mortality rate of
`those treated with active clofibrate
`(18.2%) was only slightly better than
`that of patients treated with placebo
`(19.4%). Of greater interest was the
`finding that patients with high adher-
`
`VOLUME 79, SEPTEMBER, 1997
`
`179
`
`Liquidia's Exhibit 1032
`Page 3
`
`

`

`ence rates in both groups had a dramat-
`ically lower mortality rate than poor
`adherers. That high adherence to pla-
`cebo yielded a better survival rate
`(85%)
`than low adherence (72%)
`likely reflected a difference in outlook
`and healthcare behavior between high
`and low adherers. More specifically,
`high medication adherers are also
`likely to exercise, change diet, and
`move in the direction of improved
`health more readily than low adherers.
`The expectation that one’s own behav-
`ior can and will
`lead to improved
`health is a key determinant of both
`adherence and positive clinical out-
`come.
`Caregiver-Related Factors
`The caregiver-patient relationship re-
`mains a primary determinant of treat-
`ment adherence. Patients who like and
`trust their caregiver trust the treatments
`they prescribe. Such trust is promoted
`by caregivers who are warm, friendly,
`and approachable; provide information
`and encourage communication; and al-
`low their patient a sense of control
`within the relationship and with regard
`to the treatment plan.12 The caregiver’s
`willingness to spend time with a pa-
`tient, listen to his or her concerns, and
`attempt to understand their perceptions
`and belief about
`the illness and its
`treatment
`is a positive indicator of
`commitment to changing health beliefs
`and behavior.32,33
`
`METHODS FOR MEASURING
`PATIENT ADHERENCE
`In order to identify adherence difficul-
`ties accurately or to develop effective
`adherence-promoting strategies, pa-
`tient adherence behavior must be mea-
`sured. The best measurement strategy
`for assessing adherence will be based
`on the level of precision required by
`the clinician or researcher’s goals. A
`broadly defined, flexible criterion of
`acceptable adherence may not need
`precise measurement methodology.
`When detailed and exact adherence
`data are necessary (as in research),
`however, the measurement strategies
`should be comparably precise. The
`most common measures used to assess
`
`patient adherence with asthma therapy
`are biochemical measurement, clinical
`judgement,
`self-report, medication
`measurement, pharmacy data base re-
`view, and electronic measurement.37–39
`Biochemical Measurement
`Inhaled medications are not easily de-
`tectable by biochemical assays because
`of the rapid and limited systemic ab-
`sorption of these agents. For this rea-
`son, theophylline is the only asthma
`medication adherence that
`is com-
`monly measured by biochemical assay.
`Since assays of theophylline are rou-
`tinely measured as a part of clinical
`care to determine whether a therapeu-
`tic level of
`theophylline has been
`achieved, the clinician or researcher
`can be provided with ongoing informa-
`tion about patient adherence levels.
`Biochemical measurement is the only
`adherence measurement strategy that
`provides direct confirmation of drug
`use; however,
`these measures have
`several limitations. Conclusions drawn
`from biochemical measures can be
`confounded by diet and/or other drug
`use (eg, the effect of smoking on theo-
`phylline), and these measures cannot
`be used to measure day-to-day patterns
`of adherence with therapy. Finally,
`biochemical measures can be compro-
`mised if patients deliberately, or inad-
`vertently, begin taking medications
`just before clinical samples are col-
`lected.40–46
`Clinical Judgement
`In everyday clinical care, healthcare
`providers form impressions of how
`well each patient is following the pre-
`scribed regimen. These clinical evalu-
`ations of patient adherence will shape
`the content of the patient-provider in-
`teraction, the selected therapy, and the
`follow-up plan. Several classic studies
`have shown, however, that physicians
`generally greatly overestimate the de-
`gree to which their patients comply
`with their directives. The lack of phy-
`sician accuracy in identifying patients
`with adherence difficulties has been
`attributed to a medical education focus
`that neglects communication skills and
`attention to psychosocial issues. Clin-
`
`ical judgement based on preconceived
`beliefs about the attributes of the “typ-
`ical” compliant patient are destined to
`fail. Patient characteristics such as
`race, education, sex, socioeconomic
`status, and personality have not been
`found to be reliable predictors of ad-
`herence. Physician interviewing skills
`and the qualities of the patient-pro-
`vider interaction will be more impor-
`tant in both measuring and facilitating
`adherence than stereotypical beliefs
`about adherence.47–49
`Self-Report
`Patient self-report of medication use is
`a standard measure of adherence in
`both clinical trials and behavioral in-
`tervention studies. Self-reports may be
`collected by interview, diaries, and
`questionnaires. No validated adher-
`ence-specific questionnaire is currently
`in common use, in part because most
`self-report questionnaires of adherence
`have been designed for specific stud-
`ies. Self-report measures are common
`because they are simple, inexpensive,
`and generally brief. In addition, self-
`report (particularly in the clinical set-
`ting) is the best measure for collecting
`information about patient beliefs, atti-
`tudes, and experiences with medica-
`tion regimens.
`As a quantitative measure of medi-
`cation use, self-report has been found
`to have a highly variable degree of
`accuracy. Studies by Spector et al,
`Coutts et al, and Gibson et al have
`compared asthmatic patients’ self-re-
`ports of inhaler usage with the objec-
`tive adherence data collected by an
`electronic medication monitoring de-
`vice.4,24,50 These studies have all indi-
`cated that patient self-reports of adher-
`ence
`recorded in asthma diaries
`typically overestimate adherence.
`Self-reports of adherence will be in-
`fluenced by the demand characteristics
`on the setting in which the information
`is collected. The desire to please the
`physician or investigator can lead pa-
`tients to exaggerate reports of medica-
`tion use. Physicians’ and investigators’
`skills and sensitivity in eliciting pa-
`tients’ self-reports will influence the
`reliability and usefulness of the infor-
`
`180
`
`ANNALS OF ALLERGY, ASTHMA, & IMMUNOLOGY
`
`Liquidia's Exhibit 1032
`Page 4
`
`

`

`mation they receive. While self-report
`may not be a sufficient measure of
`adherence in many settings and partic-
`ularly in research, it is probably a nec-
`essary measure in all settings. When
`carefully collected, self-reported ad-
`herence information can provide criti-
`cal insight into the nature of patients’
`problems with adherence. In addition,
`because there is no evidence to suggest
`that adhering patients will misrepre-
`sent themselves as nonadherers, self-
`report measures will identify the hon-
`est nonadherers.24,50–52
`Medication Measurement
`Counting pills, checking prescription
`refills, or weighing inhaler canisters or
`liquid medication are examples of
`medication measurement, an objective
`measure that allows researchers to in-
`fer the degree of medication adher-
`ence. This method requires recording
`the exact quantity of medication issued
`to a patient and returned by the patient
`at follow-up. Level of adherence is cal-
`culated by deriving average daily us-
`age over the monitored period. While
`medication measurement data are both
`objective and reasonably simple to col-
`lect, they are limited by several factors.
`Medication measures can be influ-
`enced by a patient’s efforts to deceive
`the investigator. Some patients may
`discard medication to appear adherent.
`Medications may be shared within
`households, particularly when family
`members are on the same medication.
`In addition, medication measures give
`no indication of the accuracy of dos-
`ages or the timing of the medication. In
`situations where patients are comfort-
`able reporting nonadherence, the pat-
`tern of medication use is not critical;
`and where the likelihood of medication
`sharing is low, medication measure-
`ment is a useful, objective, and valid
`means of assessing adherence.53,54
`Pharmacy Database-Review
`In some managed care settings, phar-
`macy databases can provide informa-
`tion on the exact regimen prescribed,
`the amount of medication dispensed,
`and the timing of refills. These data
`can be used to roughly calculate the
`
`average dose per day. In some health-
`care data management systems, pre-
`scriptions written but never filled also
`can be monitored. Dispensing data can
`also be matched with medical record
`and healthcare utilization databases to
`provide integrated analyses of the an-
`tecedents and consequences of patient
`adherence behaviors. Review of auto-
`mated pharmacy records can also al-
`low large-scale population studies of
`patient adherence with medication.
`Pharmacy database review to identify
`non-adherence has several limitations.
`First, adherence estimates can only be
`calculated for patients who exclusively
`rely on the target pharmacy system for
`all prescriptions and refills. Second,
`pharmacy data can determine when a
`prescription was filled; however,
`it
`provides no confirmation of consump-
`tion or appropriate consumption pat-
`terns. Nevertheless, as more pharmacy
`data go on-line, this adherence measur-
`ing strategy has great potential to eval-
`uate the compliance of both individu-
`als and clinical populations.55,56
`Electronic Medication Monitors
`In the past 10 years,
`the increased
`availability of computer-based tech-
`nology has introduced a new strategy
`for adherence monitoring. Electronic
`monitoring devices record and store
`the date (and for some devices, time)
`of each medication use. Devices have
`been developed to monitor medication
`adherence behaviors including, but not
`restricted to, opening a pill bottle, re-
`leasing a blister-pale pill, discharging
`inhaled medications, and releasing eye
`drops. Two electronic devices that
`have been investigated in asthma man-
`agement are the Nebulizer Chronolog
`(Medtrac Technologies,
`Inc, Lake-
`wood, CO) and the Doser (NEWMED
`Corp, Newton, MA). The Nebulizer
`Chronolog is an electronic device that
`attaches to a metered dose inhaler
`(MDI) and records each use of the
`MDI. The Nebulizer Chronolog unit
`stores the date and time for each actu-
`ation and can store up to 2000 events
`over several months. Chronolog data
`can be directly downloaded into a PC.
`The newest version of the device (un-
`
`der development) will be called the
`Medilog. The less expensive Doser de-
`vice is an electronic cap that records
`and displays daily uses of an inhaler,
`as well as remaining doses in an MDI
`canister. The Doser maintains a record
`of use for the past 30 days; however,
`this device does not record time of use.
`In addition, the Doser cannot be di-
`rectly downloaded to a PC.
`In recent years, the number of pub-
`lished studies that have used electronic
`adherence monitoring devices has dra-
`matically increased. While neither the
`Nebulizer Chronolog nor the Doser
`provides actual data on medication
`use, they provide a unique opportunity
`to investigate long-term patterns of
`presumptive adherence which were
`heretofore unavailable in such detail.
`The primary benefit of this type of
`monitoring is clear—electronic moni-
`toring methods can provide a continu-
`ous record of timing of presumptive
`doses over periods of months.
`Evaluations of adherence made by
`provider, self-report, pill counts, or
`canister weights can be inaccurate be-
`cause of recall, demand characteristics,
`deception, and provider biases. These
`methods are also insensitive to daily
`patterns of use over time. The phenom-
`enon of medication “dumping” is
`nearly impossible to detect by tradi-
`tional methods of adherence assess-
`ment, and inclusion of dumping data
`into a dose-response analysis can yield
`counterintuitive results; highly adher-
`ent subjects show poorer response than
`moderately adherent
`subjects. This
`phenomenon is likely to be present in
`any situation in which medication use
`is being monitored and should be taken
`into consideration when making med-
`ication recommendations.
`While electronic measures of adher-
`ence have
`a number of unique
`strengths, they also have a number of
`weaknesses. The cost for wide-scale
`use can be prohibitive for a small prac-
`tice and may only be feasible in a
`clinical trial setting. Additionally, fail-
`ure rate associated with the use of any
`type of electronic device may be unac-
`ceptable. The failure rate in electronic
`devices can be caused by patient mis-
`
`VOLUME 79, SEPTEMBER, 1997
`
`181
`
`Liquidia's Exhibit 1032
`Page 5
`
`

`

`use, device failure, or computer hard-
`ware/software problems. For
`these
`reasons, clinicians or
`researchers
`who use such devices must be care-
`ful
`to develop quality control pro-
`cedures that ensure the ongoing mon-
`itoring of device performance and
`validity.24,50,57,58,59
`
`IMPROVING ADHERENCE
`What Caregivers Can Do
`That
`improving treatment adherence
`can lead to better asthma control has
`been widely addressed.31,32,59,60 Recent
`discussions have increasingly recog-
`nized that the cost of introducing pro-
`grams to better educate asthmatic pa-
`tients in self-management knowledge
`and skills may be offset multifold by
`savings realized when patients require
`less emergent care and fewer hospital-
`izations.61–63 When
`programmatic
`changes include only asthma education
`classes, however,
`increased patient
`knowledge often fails to translate into
`improved disease outcomes.64,65 The
`more difficult task of changing patient
`behavior requires greater awareness of
`patients’
`individual perceptions of
`their disease and its treatment, and in-
`creased commitment from the health
`care provider to communicate with and
`teach patients. The new Guidelines for
`the Diagnosis and Management of
`Asthma7 address more comprehen-
`sively the caregiver’s role in assessing
`patient perceptions, emphasizing the
`need for a “partnership” between care-
`giver and patient to improve treatment
`adherence and disease outcome.
`The relationship between patient
`and caregiver is the single most pow-
`erful tool for changing patient health
`care behavior. Other attempts to im-
`prove adherence are unlikely to suc-
`ceed if the patient does not like and
`trust his or her doctor. Patients will not
`reveal concerns about their illness or
`reluctance about a proposed treatment
`if they believe that the caregiver is
`hurried, disinterested, or, worse yet,
`will become impatient or annoyed. The
`caregiver may say the right words—
`“Hello, how are you doing today?”—
`but simultaneously convey a nonverbal
`
`message that they are rushed and do
`not wish to be bothered by excessive
`conversation. Making direct eye con-
`tact, transmitting genuine interest in
`what the patient has to say, explaining
`all recommendations thoroughly and in
`clear language, praising good treat-
`ment adherence and problem solving,
`and expressing willingness to modify
`the treatment plan in accord with con-
`cerns expressed by the patient all en-
`hance adherence.66 Once a positive
`relationship is established, other adher-
`ence-improving changes may be nego-
`tiated, including prescribing medica-
`tions that are less costly or which avoid
`side effects concerning to the patient,
`finding reminders to help patients re-
`member when to take a medication,
`changing dosing schedules to accom-
`modate a patient’s work schedule, and
`reducing the number of medications.
`The Guidelines for the Diagnosis and
`Management of Asthma7 further rec-
`ommend that caregivers themselves
`provide patient education at the time of
`diagnosis;
`that
`they repeatedly rein-
`force the patient’s knowledge and
`skills; that they give the patient a writ-
`ten, individualized treatment plan; and
`that they remain sensitive and respon-
`sive to patient cultural and language
`differences. The Guidelines addition-
`ally provide examples of specific ques-
`tions to be used by the caregiver at
`patient visits to better address patient
`beliefs and perceptions.7
`Changing Behavior of “Difficult”
`Patients
`A subgroup of patients may be partic-
`ularly burdensome for the health care
`provider. They take extra time, create
`stress and sometimes fina