JOEL S. MINDEL
Table Of Contents
|Compliance deals with patient willingness, and ability, to take medications
as prescribed. Among the factors determining compliance are length of illness, severity
of symptoms, side effects of therapy, frequency of dosing, and patient
comprehension of the need to take drugs. The importance of these factors
rests more on logic than on scientific proof. It seems logical
that compliance would be better for a short-term symptomatic disease (e.g., bacterial conjunctivitis) than for a chronic asymptomatic one (e.g., open-angle glaucoma). Unfortunately, there are relatively few studies
of the causes of patient noncompliance; instead, most have dealt only
with its incidence. The results of those investigations of the etiology
of noncompliance have not always supported a priori logic.|
Winfield and coworkers1 investigated why patients who had been prescribed eye drops might be noncompliant:
Haynes and coworkers reviewed the evidence that decreasing the number of tablets or frequency of dosing improved compliance.2 The data were conflicting. Among the negative studies was that of Rickels and Briscoe, who found that compliance was better when neurotic patients were prescribed five instead of four pills per day.3 Gordis and associates4 and Haynes and colleagues5 found that many patients seem to take either all of their medications or none of them, irrespective of frequency of dosing. Bloch and coworkers found that compliance was increased if glaucoma patients were also receiving therapy for another chronic illness.6
Physicians are often blamed for poor compliance because their instructions to patients are inadequate. However, patients given an additional extensive consultation by a pharmacist7 or a nurse8 did not demonstrate better compliance. On the other hand, Bloch and associates6 and Vincent9 found that glaucoma patients who were compliant differed from those not compliant in their knowledge that uncontrolled glaucoma led to blindness.
Assessment of compliance presents difficulties. Objective criteria, such as the measurement of blood drug levels, usually indicate only if medication was taken within the preceding 24 hours. The patient with glaucoma who takes pilocarpine eye drops only on the morning of a doctor's appointment may be labeled as compliant because of miosis and a normal intraocular pressure, whereas the patient who forgets to take drops only on that morning may be labeled as noncompliant. Yee and associates have attempted to assess compliance indirectly.10 They have developed a monitor that records every time the patient removes the cap of an eye drop dispenser. Such devices can record not only the frequency of cap removal but also the timing. When pilocarpine was prescribed three times a day, 41% of patients missed 10% or more doses in a 3-week period; 18% of doses were at intervals of 12 hours or more, and 11% of doses were at intervals of 4 hours or less.11 Of the missed doses, more than 50% were the midday dose, and the frequency of missed doses increased the longer the time since the last clinic visit.12 These latter results suggested that reducing the frequency of eye drop administration improves compliance, as does increasing the frequency of clinic visits. Patient education, unspecified as to type, seemed to improve some patients' compliance.13 Granstrom14 used a pilocarpine electronic medication monitor during two 3-week periods. Visual fields and intraocular pressures were evaluated for 2 years before and 2 years after this monitoring in 46 patients. When pilocarpine was given three times a day, there was no significant increase in visual field loss in patients who had taken 20% or more of their drops at intervals of more than 8 hours during the monitoring period compared with those who had taken 80% or more of their drops at 8-hour intervals.
Kass and associates15,16 compared the electronic cap monitor with other means of evaluating compliance. Pilocarpine was prescribed four times a day. Intraocular pressure did not correlate with compliance as ascertained by the monitor. Patients with intraocular pressures less than 17 mmHg took 72% of their drops, whereas patients with intraocular pressures more than 25 mmHg took 68% of their drops. There were weak correlations, of 0.19 to 0.24, of the degree of compliance, as shown by the monitor, with pupillary reaction to light (69% of drops were taken by patients with reactive pupils and 80% by patients with nonreactive pupils), the physician's prediction of compliance, the patient's daily log of pilocarpine administration, the measured weight of the pilocarpine used, and the patient's self-assessment of compliance. Although subjects reported taking 97% of their drops, the monitor indicated only 76% were administered. Compliance improved during the 24 hours preceding the return visit; 88% of doses were administered, and this figure includes some patients who believed they were to omit the dose on the day of the return visit.
Alward and Wilensky17 attempted to assess compliance objectively for the oral medication acetazolamide by measuring serum carbon dioxide levels. Testing groups of inpatients allowed compliance to be controlled. Inpatients treated with acetazolamide were compared with inpatients who received no treatment. All inpatients receiving acetazolamide 500 mg or 1 g per day had serum carbon dioxide levels of 20 mEq/L or lower. Untreated control inpatients had serum carbon dioxide levels of 25 mEq/L or higher. These criteria of compliance were then applied to a group of 87 outpatients: 44% were compliant, 22% were partially compliant, and 35% were noncompliant. Surprisingly, the compliance was better in patients taking 250-mg tablets four times a day than in those taking 500-mg sustained-release capsules twice a day or 125-mg tablets four times a day.
Probably the most common objective method for evaluating compliance with oral medications has been to count the tablets remaining in the bottles. Pullar and coworkers18 formulated preparations of penicillamine (for arthritis) and glyburide (for diabetes) containing small amounts (1 or 2 mg, respectively) of phenobarbital. Phenobarbital was chosen because of its long plasma half-life and its relative lack of plasma level variation between individuals after allowances are made for weight and age. Of 161 patients with apparently good compliance by tablet count, almost half (48%) had phenobarbital concentrations lower than 90% of an age-matched volunteer group's lowest value. The volunteer group had been given correspondingly low-dose tablets of phenobarbital. Of those subgroups of patients who, by tablet count, appeared to be using excessive amounts of drug or who “forgot” to return with their bottles, the phenobarbital plasma level results were about as poor as those with apparently good compliance by tablet count. The authors concluded, “return tablet counts grossly overestimate compliance.”
Compliance can also be evaluated subjectively. Paulson and coworkers studied alcoholics taking disulfiram using an objective test (carbon disulfide breath analysis) and subjective evaluation.19 More than one third of patients claiming compliance failed the breath test. Twenty percent of those patients judged by the professional staff to be “almost certainly compliant” were not. Overall, less than 50% were compliant. In a study of antacid intake in ulcer patients,20 physicians overestimated patient compliance by 50%, and physician accuracy did not improve with greater patient familiarity.
The third form of assessment is to ask patients to admit noncompliance. Spaeth found that 31% of glaucoma clinic patients admitted being noncompliant.21 One percent blamed decreased vision from the drops. Two percent blamed other side effects. One percent stated that the medications were too expensive. The remaining 27% cited a variety of other reasons. Patients with normal vision were more likely to be noncompliant than those with impaired vision. Perhaps this should be stated in a slightly different manner: patients with normal vision were more likely to admit being noncompliant than those with impaired vision. Bloch and coworkers found that more men than women admitted being noncompliant.6 MacKean and Alkington22 found that of 169 glaucoma patients, 58% claimed compliance, 21% admitted missing an occasional dose each week, 12% admitted missing multiple doses each week but said they used at least half the medication as prescribed, 2% admitted using some but less than half the medication, and 6% admitted using none of the medications. One patient exceeded the prescribed dosage by 50%.
Kass and Becker extensively reviewed the general medical literature and found that in two thirds of the studies, 25% to 50% of outpatients failed to take their medications as directed.23 In one study, 40% of noncompliant subjects stated that they had never intended to follow instructions.
Cantrill and associates studied volunteers who responded to chronic topical corticosteroids with a high elevation of intraocular pressure.24 Subjects self-administered a series of different corticosteroids over several months. Compliance was not controlled. Dexamethasone phosphate, the first drop tested, was more potent than prednisolone acetate, the last drop tested. However, Mindel and coworkers found that when compliance was controlled, there was no difference in the pressure response produced by dexamethasone phosphate and prednisolone acetate.25 This raises the question of whether volunteer populations used in drug studies can be relied on to take their medications.
An attempt has been made by one pharmaceutical company to reduce eye drop compliance failures caused by forgetfulness. “Reminder caps” were manufactured that have a window displaying a number. The number indicates which daily dose the patient should be taking next. The window in the cap of a medication given twice daily displays a 1 or 2; that of a medication given four times daily displays a 1, 2, 3, or 4. After a drop is instilled, the patient places the cap back on the bottle and turns it until a click occurs and the next dose number appears in the window. Chang and colleagues26 used this system in bottles of a topical beta blocker, pilocarpine, dipivefrin, or any combination of these. The trial was neither masked nor controlled. The authors reported that 83% of patients found the cap easy to use. Significantly more patients (67%) claimed 100% compliance compared with the number of patients (41%) claiming 100% compliance before using the cap. Compared to before using the cap, there was an overall additional reduction in intraocular pressure of 0.8 mmHg after the reminder cap was introduced. The subset of patients claiming increased compliance had a reduction in intraocular pressure of 1.7 mmHg compared with their “pre-cap” baseline.
Although the problem of noncompliance is frustrating, there is one beneficial aspect. The toxic effects of drugs are avoided. Sarin and coworkers found that psychiatric outpatients receiving large doses of chlorpromazine, averaging 160 mg/day for 9 to 12 years, failed to develop the toxic eye findings associated with inpatient therapy.27 They pointed out that noncompliance was a likely explanation.
4. Gordis L, Markowitz M, Lilienfield AM: Studies in the epidemiology and preventability of rheumatic fever: IV. A quantitative determination of compliance on oral penicillin prophylaxis. Pediatrics 43:173, 1969