Barry Blackwell: Pioneers and Controversies in Psychopharmacology 
Chapter 22: Compliance and the Therapeutic Alliance 


       The reader may wonder why the final chapter is on this topic.  For two reasons: first, even the most effective and safest drug is useless when a patient declines to take it. Full stop! This is a frustrating truism as old as Hippocrates’ aphorisms: “keep watch also on the faults of the patients, which often make them lie about taking of things prescribed” (Wright 1993) and “a regimen too strict and unsubstantial is always dangerous … when it does not agree with the patient” (Coar 1882).  

       These are frustrating aphorisms that must have confronted John Cade, Jean Delay and all the pioneers in the early days of psychotropic drugs when psychotic and delusional patients were offered treatments known to be effective and they declined to trust or take them.  

       Secondly, compliance has been a career-long interest, one among many “non-drug factors” that influence, enhance, blur or confound drug therapy. I published my first article on the topic after an invitation by the New England Journal of Medicine (Blackwell 1973) and the final book of my career was published almost a quarter century later with the same title as this chapter (Blackwell 1997).  

       Awareness and interest was initiated by McMaster University in Toronto when the Medical School convened two International Conferences on Compliance in 1974 and 1977 to which I contributed. As a consequence, the term “patient compliance” appeared for the first time in 1975 when the Index Medicus used it to replace the term “Patient drop out” – almost a quarter century after psychotropic medications began to appear.  

General Concerns and Consensus 

      By the time of the millennium more than 12,000 articles on the topic had appeared, continuing at the rate of 900 annually. The 34 authors contributing to my book revealed the depth and breadth of interest. They included a consumer, social worker, psychologist, pharmacist, educator, law professor and administrator, as well as family members, physicians, community support workers and psychiatrists of different stripes: forensic, geriatric, community, addictions and behavioral health.  

       The problem of compliance is ubiquitous across all areas – more than 50 different diseases or conditions – the most common being pediatrics and psychiatry followed by hypertension, smoking, Asthma/COPD and diabetes. Its diverse manifestations may include failed appointments, reluctance to initiate treatment or follow recommendations involving diet, lifestyle or prescriptions.  Drugs may be taken for the wrong conditions, at the incorrect time or dosage, in the wrong dose or not at all. The annual cost of non-compliance was estimated to be in excess of a billion dollars annually in hospital or nursing home admissions, lost work productivity, premature deaths and outpatient treatment costs (Reston 1992) but must certainly be far more today.  

        Measurement is complicated – the problem tends to disappear once it is being observed; simple measures are not accurate and accurate measures are not simple. How much compliance is necessary for a good outcome varies – usually 80% for hypertension, a level at which poor dietary control and weight gain may occur. Even at 100% high blood pressure may persist if a person is overweight, under stress or absorbs the drug slowly but metabolizes it quickly.  

       An early study found that only one in five studies had an adequate research design (Haynes et al. 1979) and at least half failed to yield a relationship between compliance and outcome measures (Blackwell 1989). Generalization is poor; factors that influence clinic attendance differ significantly from those that influence medication taking (Melnikow and Kiefe 1994).  

        Nevertheless, there are some generalizations and limited consensus from this large field of enquiry. Roughly 25% of inpatients and 50% of outpatients are non-compliant with medication and interventions are more successful compared to changes in lifestyle. On the positive side compliance is more likely if the patient’s treatment expectations are met; if they are supervised or have continuity of care and the disease is a serious one to which the person feels susceptible. Multimodal and sustained interventions are more effective than single or sporadic ones. Involvement of family or friends often makes a significant contribution.  

       Poor compliance is fostered by lengthy treatment, asymptomatic disorders, complicated regimens, side effects, social stress, isolation or alcoholism and lack of insight in psychotic disorders (Amador et al. 1993). 

Factors Affecting the Prescribing Psychiatrist. 

       To forge an effective treatment alliance requires knowledge of both the generic demands that illness imposes on the patient and the physician, as well as issues associated with persistent mental illness, psychotropic medication, the treatment regimen and the patient’s symptoms. Above all, an alliance implies a deep level of intuitive understanding of the patient’s condition, acceptance and mutual respect.  

The Physician 

      Attempts to understand compliance have focused almost entirely on the attitudes, beliefs and behaviors of the patient and hardly at all on the physician. Paradoxically, research reveals that both physicians and patients seriously overestimate the extent to which people do what the doctor tells them. 

      The doctor’s beliefs and behaviors stem directly from the unique stress of the physician’s role: making difficult decisions in ambiguous situations, the responsibility for finding a cure and taking control.  This propensity, useful in emergencies, is learned in medical school and residency but often carries over into day-to-day practice where “caring” exceeds “curing,” despite educational attempts to stifle it. Often the patient’s anxiety in the face of illness and the doctor’s “take charge” demeanor quells assertiveness in favor of a passive rather than a partnership role.  

       Over time and with experience psychiatrists and many physicians come to realize that caring for persons with severe or intermittent mental illness requires much more negotiation and participation before providing answers. 

The Patient 

         All long-term illnesses impose generic burdens on those who suffer. These challenge the core human needs to feel in control, autonomous, intact and connected to others. Severe illness arouses feelings of ambiguity, loss of integrity, dependence on others, often stigma and isolation. Each individual’s way of dealing with these issues will reflect their coping capacity, personality structure and experiences. 

        Just like physicians, patients possess unique beliefs, attitudes and experiences concerning their illness. These may include its cause, the appropriateness or otherwise of a particular treatment and the likely outcome. Mutual discussion and negotiated agreement of these issues lays a solid foundation for a productive alliance. For example, people with strong beliefs in the spiritual or psychological origins of illness are likely to prefer talking to taking medications. Sometimes with chronic illness, and almost always with homeless persons, an individual may have had bruising experiences with uncaring or coercive treatment which makes them slow to accept advice or help. Many may share a belief that the body sometimes “needs a rest” from continuous medication and they should take “drug holidays,” an attitude that frustrates prophylaxis in recurring conditions.  

The Illness 

       Each major disorder’s symptoms are grafted onto a person’s predisposition and together these can interfere with the alliance or treatment plan. Poor concentration, impaired memory, thought disorder or delusions can interfere with information exchange or agreement. Depression, apathy or the negative symptoms of schizophrenia can diminish persistence and motivation necessary for prolonged treatment. Lack of awareness or acceptance of illness, “insight” is sometimes a biological manifestation, especially in acute mania of bipolar disorder (Amador et al. 1993) but also, at times, a psychological reaction to the stigma of being labelled with a mental illness and a struggle to retain respect and self-worth. 

       Of great practical significance are the difficulties encountered by patients with symptom sensitivity (somatization), bodily concern (hypochondriasis) or panic disorder where episodic autonomic over-arousal affirms they are “allergic” to allopathic medications. They may be attracted to homeopathic or alternative therapies that are ineffectual but not necessarily harmful.  

       Co-morbid substance abuse is a risk factor in all forms of poor compliance but especially in psychiatric illness because alcohol or drugs may interfere with the therapeutic effects of treatment or may be used as self-treatment to stifle psychotic symptoms or depression.  

       In contrast to this bleak litany, major psychiatric disorders have features that can facilitate compliance. Unlike hypertension, glaucoma and diabetes, which are “silent” conditions, mental illness produces symptoms the patient may be motivated to suppress. Because these disorders are intermittent and recurring it also creates an opportunity to learn the linkage between taking medication and symptom relief.   

       Finally, the onset and prevalence of major psychiatric disorders often occur in early to mid- life when cognitive skills and social connections are better preserved than with the degenerative physical and memory disorders associated with ageing.  

Psychotropic Medications 

       Psychotropic drugs and their target organ possess properties that complicate compliance. The brain is well protected and for drugs to penetrate the blood-brain barrier requires significant amounts to be absorbed before they are metabolized and excreted. The receptors with which these drugs interact may exist elsewhere in the body, not just the brain, often in the peripheral nervous system or the gut, which develops from the same part of the embryo as the brain. This ensures that unwanted side effects are common, often effecting cardiovascular, sexual or gastrointestinal function.  

       Changes in brain chemistry and function often occur slowly due to up or down regulation of receptors. Most therapeutic effects are delayed while many side effects occur swiftly. This asynchrony is counterintuitive and an invitation to poor compliance because people are accustomed to the immediate action on the brain of alcohol, caffeine, nicotine or drugs of abuse. Paradoxically, compliance may be a manifestation of dependence on drugs with a short half-life when missing a dose initiates withdrawal effects. For this reason, Alprazolam (Xanax) became known as the “American Express pill; don’t leave home without it.” 

The Treatment Regimen 

       Compliance is facilitated when the regimen fits the lifestyle and convenience of the patient. It is not usually compromised unless the patient is taking multiple medications or one medication more than three times a day (Blackwell 1979). Historically, medicines were taken that way with meals, still the principal cue. Nowadays, plastic pill boxes have become a ubiquitous compliance adjunct, placed in a prominent place in plain view as a perpetual reminder. 

       Compliance is facilitated when medication benefits and side effects are clearly defined; the former include restful sleep, increased energy, clear thinking and an absence of voices or alien ideas. Side effects are a major deterrent especially if they are unexpected, severe and sudden or interfere with normal functions. Linkage between medication benefits and side effects can often be clarified by having a person keep daily records, especially patients who are symptom sensitive (somatization or hypochondriasis). 

The Support System 

       Another concerned person is among the most potent facilitators for compliance – a therapist, relative, friend, case manager or payee. Critical or intrusive involvement may be counter-productive and provoke psychotic episodes in vulnerable individuals (Vaughn and Leff 1977).  

The Alliance 

       A physician’s need for control and cure may rub up against the patient’s desire for autonomy. If the doctor evokes fear or uses coercion these can arouse resistance or regression. In contrast, empowerment and personal control reinforce belief and participation in treatment.  

       A productive alliance requires the physician inquire about the patient’s attitudes, beliefs and knowledge concerning their disorder and treatment options. If these differ from the physician’s ideas negotiation may be necessary. For example, belief in megavitamin therapy can be accommodated if the patient agrees to take traditional medication. A sensitive awareness of a person’s dynamics may avoid stirring up resistance to an overly parenteral or didactic approach which itself may encourage regression and passive adoption of the sick role. The patient’s ability and willingness to become active in the alliance can also be shaped by the phase and severity of illness. In acute episodes psychotic symptoms, poor insight or impulse control are impediments; later on, apathy, poor morale or negative symptoms may interfere.  

       Long term success in chronic or recurring disorders is shaped by the patient’s willingness to learn how medications modify symptoms or side effects and the cost-benefit ratio of different treatment regimens.  

       Finally, changes in the psychiatrist’s role in recent years due to insurance company reimbursement policies have complicated the alliance; 15-minute med checks with the capitation of therapy and prescribing to different disciplines have created problems for both patients and prescribers.  

        The above aspects of the alliance are illustrated in the following brief clinical vignettes. 

Reconciling schizoaffective disorder, HIV infection and cocaine abuse in a young African American male 

       Randolph is a 30-year-old intravenous drug abuser with schizoaffective disorder who recently became positive for HIV. During a recent hospital admission for medical treatment the nurses felt he was paranoid and depressed. His urine was positive for cocaine and he rejected offers of medication for his mental illness.  Questioned at the bedside he admitted to using cocaine to stifle feelings of “disgust.” Randolph also agreed that after each brief euphoric interlude he was quickly back in the real world, feeling even more suspicious and disgusted. Our conversation began with a detailed analysis of why Randolph used cocaine, what it did for him and how he thought it worked. Based on his experience, Randolph agreed that drugs could alter mood and behavior but at some cost. I suggested he might consider an alternative in the form of prescribed medications. I pointed out that unlike cocaine these worked slowly, that benefit was gradual but that, unlike cocaine, the benefits persisted. If he was willing to try an experiment chances were he would feel less suspicious and disgusted. As a bonus he might no longer crave cocaine. Randolph agreed it was worth a try. After discharge he came to our homeless clinic and was started on an antipsychotic and antidepressant. Two weeks later he was pleased and surprised to report that his voices were fading away, he was sleeping soundly and feeling much less disgusted with life. His urine was clean.  

Resolving recidivism in a midlife single Caucasian woman with bipolar disorder 

       Agnes is a 42-year-old single female well known throughout our city for frequent admissions to several hospitals over many years. She usually arrived accompanied by the police in an acute manic episode, combative, provocative and totally lacking any capacity to view herself as ill or in need of treatment. Although highly intelligent, Agnes was unable to hold a job and received Medicaid with health benefits. She was unable to find a private psychiatrist willing to treat her and was unpopular with the inpatient staffs who viewed her behavior as “borderline.” 

       Our first few months were spent finding more about each other, during which time Agnes was admitted twice to our teaching unit. Agnes was willing to take medication I prescribed but detested the way mood stabilizers stifled her creativity; she experimented continuously with various forms of meditation and marginal religious cults giving herself frequent “drug holidays.” 

       In our initial dialogues each of us attempted to convert the other. I quickly learned how sensitive Agnes was to control issues; her father had been a benevolent tyrant and her mother was critical and unloving. Agnes found her experiences with the police and hospital staff humiliating. I encouraged her ongoing search for spiritual meaning and non-drug methods of thought control but I also maintained that these were compatible with taking medication if she was willing to explore the tradeoff between a benefit she valued – staying out of hospital – and side effects she disliked – slowing of thoughts. Together we agreed on a target: to persist with medication in a sustained way until Agnes had been out of hospital for longer than any other interlude; about 11 months.  

Acknowledging and healing the damage done by 15-minute med checks to the treatment alliance 

       In the summer of 1996 my task as chair of the Wisconsin Psychiatric Association’s Committee on Liaison with Community Agencies led me to conduct a survey of the seven major organizations. It was triggered by deep concern among our members about the constraints managed care and insurance companies were placing on physicians providing psychotherapy which could be less expensively provided by social workers and psychologists. This segregated medication from therapy and gave birth to the 15-minute med check. I conducted seven-hour long focus groups with consumers convened by the agencies. The results and my response were published in the Society’s fall newsletter (Blackwell 1996). 

                                          Consumers’ verbatim comments were as follows: 

There has to be something to talk about other than medication. 

You’re treated like a thing, by a technician. 

In the door, out the door, because the clock’s ticking. 

People shouldn’t be giving medicines unless they know the person. 

There are two types of psychiatrists, the talking one and the one that does refills – we feel shortchanged. 

All they ask real quick how you are doing – by the time I leave I remember all kinds of things. 


       My colleagues’ reflexive response to this clear-cut consensus was understandable but unhelpful; if they were only paid more they could do a better job with the additional time.  My own published comments were as follows: 

“But what is a med check? What distinguishes a good from a bad one? Although I have been a teacher for thirty years and have a major interest in compliance I have not asked myself these questions and I cannot remember teaching a resident the necessary skills.” 

       At the most basic level, a med check (defined by the insurance industry) is 15 minutes of a psychiatrist’s time billed at more than 15 times the minimum wage. Do we give value for money?  Some psychiatrists seem to feel that such a short time is inherently useless. While it is a lot less than our customary 50-minute hour, it is all primary care providers schedule for a follow up office visit.  That is about the average time I spend with the dozen or so people I meet every Thursday evening in our Health Care for Homeless Clinic. As in primary care, our knowledge of each other is cumulative and sequential – three hours of acquaintance annually, on average.  

       Seeing a lot of people in a short time is a stress on anyone’s memory. I always refresh my recall from personal details in the chart so that the new visit is linked to the previous one. I read the chart and then go to the waiting room to greet the patient by name. As we sit down I ask, “How did your daughter’s wedding go?” “How was your vacation?” Has your leg healed up?” “Did you finish painting the kitchen?” 

      From these specific questions I move on to the general issue of how life is proceeding; job, marriage, children, leisure? Any new or unusual events? Once I learn these things I already have an idea of what the medication is or is not doing. So now I am more direct – after all this is a med check. There are several issues to touch on but the amount of time for each varies with the phase of treatment. Early on I focus on knowledge of the medication, its name, dosage, regimen, and likely duration of treatment. I emphasize the delayed onset till improvement and common side effects that occur early. I advise the patient about the use of alcohol, what to do if a dose is forgotten and explore the best way to adjust treatment to lifestyle. With mood stabilizers I explain target blood levels.  

       I pay particular attention to compliance, making forgetfulness easy to admit. A final question is about the extent to which quality of life is improved. I try to interpret drug action in terms of everyday events. Does it improve enjoyment, benefit sleep, diminish voices, decrease energy, impair sexual function or cause weight gain? A final brief inquiry is about contacts with other therapists or physicians. This is to encourage collaboration and avoid splitting. It makes me aware of whatever psychosocial issues or difficulties are being addressed and alerts me to any need to communicate with primary care physicians about the medications they are prescribing. 

      Conceptually, I distinguish three components to this approach: first, I establish a personal relationship; second, medication is offered to maximize quality of life; and lastly, its use is integrated with other psychosocial or medical interventions 

        If psychiatrists agreed on the components of a med check it would facilitate training, measurement, modification and improvement. The false dichotomy between talking and prescribing would vanish. Communicating these goals to consumers would create realistic expectations about what can be done in 15 minutes and satisfaction might be improved. A name change might help. How about Continuous Integrated Medication Management? Sounds better than “Med Check.”  


Amador XG, Strauss DH, Yale SA, Flann MM, Endicott J, Gorman JM. Assessment of insight in psychosis. Am J Psychiatry. 1993 Jun;150(6):873-9. 

Blackwell B. Patient compliance. N Engl J Med 1973; 289:249-252. 

Blackwell B. The Drug Regimen. In: Haynes RB, Sackett DL. Compliance in Health Care. Johns Hopkins University Press, 1979, pp. 144. 

Blackwell B. Compliance – Measurement and Intervention. Current Opinion in Psychiatry. 1989, 2: 287-789. 

Blackwell B. From Compliance to Alliance: A Quarter Century of Research in Treatment Compliance and the Therapeutic Alliance (Ed, Blackwell B) Amsterdam; Overseas Publishers Association, Harwood Academic Publishers. 1997; 1-16. 

Blackwell B. Anatomy of a Med-Check. WisconsinPsychiatrist, Fall 1996. 

Coar T. The Aphorisms of Hippocrates; With a Translation into Latin, and English. 1822. London, Longman.  

Melinkow J, Kief C. Patient compliance and medical research; issues and methodology. J Gen Intern Med. 1994 Feb;9(2):96-105. 

Reston VA. Emerging issues in pharmaceutical cost containment. National Pharmaceutical Council. 1992, 2: 1-16. 

Vaughn CE, Leff JP. The influence of family and social factors on the course of psychiatric illness. A comparison of schizophrenic and depressed neurotic patients. Br J Psychiatry. 1976 Aug; 129:125-37. 

Wright EC. Non-compliance--or how many aunts has Matilda? Lancet. 1993 Oct 9;342(8876):909-13. 


August 23, 2018