AlignMap

The Presentation

Compliance With Capecitabine Therapy Very High Among Swiss Cancer Patients: Presented at ASCO-GI ¹  summarizes a presentation of study results² made on January 17 at the American Society of Clinical Oncology’s 6th Gastrointestinal Cancers Symposium (ASCO-GI), cosponsored by the American Gastrointestinal Association Foundation, the American Society for Radiation Oncology, and the Society of Surgical Oncology.

The findings are encouraging – but there is a caveat. Pertinent excerpts follow:

Patients who were prescribed oral capecitabine for treatment of breast or gastrointestinal (GI) cancer appear to be highly compliant in taking the oral medication, according to a study conducted in Switzerland.

… For the study, patients receiving capecitabine either as monotherapy or in combination with other chemotherapeutic agents recorded their daily capecitabine intake and any adverse effects on a diary. After completion of therapy for a maximum of 8 cycles, the data were transferred to a questionnaire in which the reasons for discontinuation were also collected.

…  Of the overall cohort, 91% took capecitabine as prescribed for the entire course of treatment. Reasons for interrupting therapy included forgetting to take treatment (56%), adverse effects (25%), and misunderstanding instructions (19%).

… Dr. Winterhalder said that the 16 patients who did make compliance errors included 9 instances in which the patients forgot to take the medicine. He said that despite the impressive compliance figures seen in the study, there are ways to improve the compliance further.

“Patient management systems such as patient diaries may further improve compliance and adherence with treatment,” he said during his poster presentation. “Compliance may be further improved by educating patients about how to recognise and manage treatment toxicities.”

Commentary Or What’s Wrong With This Premise?

Maybe I’m missing something, but I think this study declares that using a medication diary enhances compliance – based on compliance rates measured by the patients’ entries in a medication diary.

And, as far as I can determine, there is no control group for comparison.

Either the summary of this study is incomplete or the quality of the evidence must be considered suspect.

1. Compliance With Capecitabine Therapy Very High Among Swiss Cancer Patients: Presented at ASCO-GI By Ed Susman. Doctor’s Guide. January 19, 2009↩
2. Presentation title: Compliance With Capecitabine: Findings From a Prospective Cohort Analysis. Abstract 420↩

From Wired’s list of 10  Top Technology Breakthroughs of 2008

9. Edible Chips

Grandma’s pillbox with the days of the week neatly marked is set to go high tech. Tiny edible chips will replace the organizer, tracking when patients take their pills (or don’t) and monitoring the effects of the drugs they’re taking. Proteus, a Redwood City, California, company, has created tiny chips out of silicon grains that, once swallowed, activate in the stomach. The chips send a signal to an external patch that monitors vital parameters such as heart rate, temperature, state of wakefulness or body angle.

The data is then sent to an online repository or a cellphone for the physician and the patient to track. Proteus says its chips can keep score of how patients are responding to the medication. That may be just the beginning, as the chips could improve drug delivery and even insert other kinds of health monitors inside the body. Now doctors may have a better answer to a common patient complaint — they will know exactly how it feels.

Outlook: If proven in clinical trials, edible chips could let physicians look into a patient’s system in a way that could change how medicine is prescribed and how we take the drugs.

Commentary

While I am less confident than the denizens of Wired that the edible chips will automatically mark a revolution in health care, I do believe they could be a key tool in researching and confirming patient compliance.

At the least, they should be the new gold standard for tracking medication adherence.

Credit Due Department: Image from Businessweek

Because of the response to the  previous post,  Clinicians Can Choose How, Not If, They Influence Patient Compliance, I am posting references to two recent publications that address issues raised in that entry.

Cyberchondria

Microsoft has published the results of a study of cyberchondria, the behavior of individuals who use the Internet to gather information on health or healthcare. The Microsoft study tallies data from health-related Web searches on popular search engines (a 40 million page anonymized sample) as well as a survey of 500 Microsoft employees about their own health-search experiences.

The New York Times summarized the report’s findings:

The study suggests that self-diagnosis by search engine frequently leads Web searchers to conclude the worst about what ails them. … Mr. Horvitz, an artificial intelligence researcher at Microsoft Research, said many people treated search engines as if they could answer questions like a human expert. “People tend to look at just the first couple results,” Mr. Horvitz said. “If they find ‘brain tumor’ or ‘A.L.S.,’ that’s their launching point.”  … They found that Web searches for things like headache and chest pain were just as likely or more likely to lead people to pages describing serious conditions as benign ones, even though the serious illnesses are much more rare.  For example, there were just as many results that linked headaches with brain tumors as with caffeine withdrawal, although the chance of having a brain tumor is infinitesimally small.

The study can be downloaded at Microsoft Cyberchondria Report

Survey Of Placebo Prescribing

Last month, BMJ published the results of a survey of the habits of American physicians re the prescription of placebos.The abstract follows:

Prescribing “placebo treatments”: results of national survey of US internists and rheumatologists

Objective To describe the attitudes and behaviours regarding placebo treatments, defined as a treatment whose benefits derive from positive patient expectations and not from the physiological mechanism of the treatment itself.

Design Cross sectional mailed survey.

Setting Physicians’ clinical practices.

Participants 1200 practising internists and rheumatologists in the United States.

Main outcome measures Investigators measured physicians’ self reported behaviours and attitudes concerning the use of placebo treatments, including measures of whether they would use or had recommended a “placebo treatment,” their ethical judgments about the practice, what they recommended as placebo treatments, and how they typically communicate with patients about the practice.

Results 679 physicians (57%) responded to the survey. About half of the surveyed internists and rheumatologists reported prescribing placebo treatments on a regular basis (46-58%, depending on how the question was phrased). Most physicians (399, 62%) believed the practice to be ethically permissible. Few reported using saline (18, 3%) or sugar pills (12, 2%) as placebo treatments, while large proportions reported using over the counter analgesics (267, 41%) and vitamins (243, 38%) as placebo treatments within the past year. A small but notable proportion of physicians reported using antibiotics (86, 13%) and sedatives (86, 13%) as placebo treatments during the same period. Furthermore, physicians who use placebo treatments most commonly describe them to patients as a potentially beneficial medicine or treatment not typically used for their condition (241, 68%); only rarely do they explicitly describe them as placebos (18, 5%).

Conclusions Prescribing placebo treatments seems to be common and is viewed as ethically permissible among the surveyed US internists and rheumatologists. Vitamins and over the counter analgesics are the most commonly used treatments. Physicians might not be fully transparent with their patients about the use of placebos and might have mixed motivations for recommending such treatments.

The citation and links to the abstract and full text (available without charge) follow:

Tilburt, J. C, Emanuel, E. J, Kaptchuk, T. J, Curlin, F. A, Miller, F. G (2008). Prescribing “placebo treatments”: results of national survey of US internists and rheumatologists. BMJ 337: a1938-a1938  [Abstract] [Full Text]

Source: Cohort study examining tamoxifen adherence and its relationship to mortality in women with breast cancer. C McCowan, J Shearer, P T Donnan, J A Dewar, M Crilly, A M Thompson and T P Fahey. Br J Cancer advance online publication, November 4, 2008; doi:10.1038. Link To Abstract

Abstract

Increasing duration of tamoxifen therapy improves survival in women with breast cancer but the impact of adherence to tamoxifen on mortality is unclear. This study investigated whether women prescribed tamoxifen after surgery for breast cancer adhered to their prescription and whether adherence influenced survival. A retrospective cohort study of all women with incident breast cancer in the Tayside region of Scotland between 1993 and 2002 was linked to encashed prescription records to calculate adherence to tamoxifen. Survival analysis was used to determine the effect of adherence on all-cause mortality. In all 2080 patients formed the study cohort with 1633 (79%) prescribed tamoxifen. The median duration of use was 2.42 years (IQR=1.04–4.89 years). Longer duration was associated with better survival but this varied over time. The hazard ratio for mortality in relation to duration at 2.4 years was 0.85, 95% CI=0.83–0.87. Median adherence to tamoxifen was 93% (interquartile range=84–100%). Adherence <80% was associated with poorer survival, hazard ratio 1.10, 95% CI=1.001–1.21. Persistence with tamoxifen was modest with only 49% continuing therapy for 5 years of those followed up for 5 years or more. Increased duration of tamoxifen reduces the risk of death, although one in two women do not complete the recommended 5-year course of treatment. A significant proportion of women have low adherence to tamoxifen and are at increased risk of death.

Commentary

Other than the specifics – i.e., which medication,  which diagnosis, which patient cohort, etc – there is little new information offered here.  The primary messages are clear:

• Decreased adherence (to tamoxifen in this case) is associated with decreased survival.
• Ongoing use of tamoxifen reduces the risk of death, yet only 49% of patients continue therapy for 5 years.

I had not planned to post this week, but this reiteration of the all too familiar patient noncompliance motif is just too damn sad to ignore.

One of my patient compliance alerts this morning linked to Drug Compliance: Barriers to Care at Endo Blog where I found the attractive chart atop this post along with a discussion of the data it displays. Those findings are summarized in this excerpt:

From my reading of the post, it appears that the focus is on extracting from the chart (and the article whence the chart originated) practical recommendations for diabetic patients to enhance their ability to follow their treatment regimens. And from that perspective, the post is on target. The author advises, for example,

I’ve issued similar suggestions to my own patients, after learning (the hard way, of course) not to assume that patients were not born with the knowledge that orders for prescription refills could not always be issued immediately nor could the pharmacy always immediately produce a bottle of pills.

So far, so good. But I do want to use this post as an example of the risks of summarizing a study’s findings too concisely or, as I suspect is the case here, to present that summary in a public forum with only one segment of the potential audience in mind.

The Alternative Interpretation

On tracking down the source of the chart from the legend on its lower left corner, I recognized the article, Barriers to Medication Adherence in Poorly Controlled Diabetes Mellitus,¹ as one I had previously read and, in fact, had referenced one conclusion, the link between difficulty reading the text of the prescription and poor control of diabetes as indicated by the A1c biological marker, in my own post, Check The Fine Print For Noncompliance – Part 1.

Excerpts from the study’s abstract follow:

The problem lies not in what was written about the study but in what was not written.

Specifically, the following information wasn’t available to those who read only the post about the study:

1. The patient population was selected in part because of their poor control of their diabetes, as signaled by A1c values of 9% and above
2. Data re adherence and the challenges to adherence were collected exclusively by self-report
3. As noted in the abstract, “Self-reported adherence was not related to A1c control,” but “Taking more than 2 doses of DM medication daily and difficulty reading the DM medication prescription label were significantly associated with higher hemoglobin A1c.”

Now, the absence of that information has little or no impact on the advice offered to patients in the post. And the phrase, “according to patients, … ,” does indicate the source of the information.

Nonetheless, that declaration of results, while hardly egregious, is problematic, as can be seen by comparing the statement as written to a more complete version of the information.

As written:

More complete version:

My contention is that those two versions may have significantly different impacts, at least on certain readers.

And that is the point: public blogs are – well, public. Posts can be read by, among others, patients, clinicians, reporters, elected officials hoping to find justification for public policy changes, students writing doctoral papers, nurses from Africa caring for HIV patients who deny they are infected, lawyers working the compliance angles on behalf of their clients, colleagues with points of view congruent with the content, colleges with opposing points of view, marketing folks from pharmaceutical companies, and other bloggers.

That is only a partial list of those who have contacted me after reading something at AlignMap.com; that list is also the reason I write – and urge other bloggers to write – with the assumption that their readers will come to their posts with an extraordinarily wide variety of experience with and knowledge of the topic and with an even wider range of motivations.


Footnotes

1. Peggy Soule Odegard and Shelly L. Gray, Barriers to Medication Adherence in Poorly Controlled Diabetes Mellitus, The Diabetes Educator, Vol. 34, No. 4, 692-697, 2008↩

Alternative Perspectives On Patient Compliance

We Interrupt This Rant, …
As ongoing readers know, recent AlignMap posts have been a series of jeremiads lamenting both (1) the repetitive nature of patient compliance research, programs, and theoretical thinking and the resultant paucity of advances in the field and (2) the questionable value of  recurrent skirmishes over details such as the most appropriate appellation du jour for the system currently known as patient compliance, a battle which strikes me as the equivalent of a bidding war for naming rights to the Titanic breaking out just after the collision with the iceberg.

Well, to invoke the astoundingly convenient Monty Python pseudo-segue, now for something completely different, i.e., an atypically positive post suggesting a pragmatic means of expanding the conventional knowledge base of patient compliance. While that appropriately modest goal falls short of a universal panacea for treatment failure, the redemption of men’s souls, or the establishment of cosmic justice, it’s not a bad way to start the week.¹

Patient Compliance Information Source Alternatives:
We Are Not Alone

The key to unlocking a wealth of information and thoughtful research with direct and inferential links to treatment adherence is the willingness to consider the possibility that the two-part iconoclastic hypothesis presented a few lines below may, however incredible  it may seem, be valid.

Before revealing this fundamental reshaping of the intellectual firmament, authorial responsibility  dictates that I issue certain caveats. Those easily shocked, those with sensitive temperaments, and those diagnosed with high anxiety, severe cardiac conditions, or other disorders known to be  exacerbated by strong  emotional or intellectual challenges may wish to confer with their personal physicians before continuing. Medications, if appropriate to the situation, should be at hand. Ladies and older gentlemen, even those in superb health, should be seated or recumbent upon reading the remainder of this post. Those who feel they cannot tolerate further chaos in their lives at this point should cease reading no later than the end of this paragraph.  Knowing ones own limitations is a strength, not a weakness. The  official AlignMap Blog position holds that  there is no shame in dropping out now rather than risk ones wellbeing.

Those intrepid souls determined to pursue this idea should now prepare themselves.

Precursor Principles For Expanding The Patient Compliance Model

Principle 1. Patients are not exclusively patients. Reliable evidence has begun to accumulate, for example, that some individuals, despite meeting rigid criteria identifying them as “patient,”  also hold  jobs, sometimes devoting 40 hours or more a week to their occupational roles. Others are now known to operate as parents, grandparents, brothers, sisters, friends, partners, and a myriad of other roles. Rumors have even arisen that many patients have strong positive and negative feelings toward others that seem to have nothing to do with health or healthcare. There have been confirmed sightings of patients functioning simultaneously in several social, cultural, and spiritual spheres independent of their medical treatment status. Further, many patient brazenly and casually admit to these non-clinical identities and invest considerable psychological resources in them. At a minimum, these observations cast doubt on prevailing Patient Theory which holds that patients, when not in the presence of a clinician or in the act of executing a prescribed treatment, are maintained in a state of suspended animation until awakened for their next clinic appointment or medication dose.

Principle 2. The processes that culminate in Patient Compliance or Noncompliance do not operate exclusively in matters of health and healthcare. In fact, the manner in which a patient responds to treatment recommendations from a clinician and the extent to which that patient follows those treatment recommendations may be similar to the manner in which that person responds to and follows recommendations from a lawyer, a broker, a business consultant,  a teacher, a military superior, a friend, a mechanically derived algorithm, …  – even if those  recommendations have no direct implications for healthcare.

Heady stuff, eh?

It’s a lot to digest, but there is a payoff. Because of the extensive data, research, and literature available about how people respond to and follow those non-healthcare recommendations (often called “advice” in the non-medical world), these metaphysical musings transform into something real – and something immediately useful. In the fields of psychology (in this case, that portion of psychology not directly linked to medicine), sociology, economics, political science, education, business, and market research, among others, a plethora of data, interpretations, studies, and reports exist under topical headings such as decision-making, the spread of ideas, purchase resistance, learning processes, behavioral influences, … .

And, even better, most of that material is not a rehashing of the medical literature on patient compliance, but, in fact, may offer  perspectives that are unique from yet could be applicable to clinical adherence.

Serendipitously, an example is at hand.

The Impact Of Emotion On Patient Compliance

Source: Feeling the Love (or Anger): How Emotions Can Distort the Way We Respond to Advice Knowledge@Wharton, October 1, 2008

Knowledge@Wharton is the online newsletter of the Wharton School of the University of Pennsylvania. Wharton is, of course, an eminent business school and the newsletter is congruently oriented.

I’m not covering the article in depth. Instead, I will present excerpts to give a flavor of the entire piece, point out some specific elements I think are significant to those of us invested in understanding patient compliance, and, finally, invite the reader to review the original essay itself along with the relevant research on which the article is based. Both the article and the essay are available on the same Knowledge@Wharton web page.

Here’s a piece of advice: Don’t read this story if you have just had a fight with your spouse or a co-worker. You will probably ignore it, despite its grounding in solid academic research. At least that’s what Maurice Schweitzer, a Wharton professor of operations and information management, would most likely suggest. In a recent paper written with Francesca Gino of Carnegie Mellon University, he shows that emotions not only influence people’s receptiveness to advice but they do so even when the emotions have no link to the advice or the adviser.

“We focus on incidental emotions, emotions triggered by a prior experience that is irrelevant to the current situation,” the two scholars note in their paper, titled “Blinded by Anger or Feeling the Love: How Emotions Influence Advice Taking.” “We find that people who feel incidental gratitude are more trusting and more receptive to advice than are people in a neutral emotional state, and that people in a neutral state are more trusting and more receptive to advice than are people who feel incidental anger.”

… until recently, economic analysis has taken as its premise the idea that, when it comes to dollars and cents, people can wall off their emotions. “Classical economics is predicated on this rational-man idea and also on the idea that mistakes will get extinguished by the market,” Schweitzer says.

But Schweitzer and Gino’s research suggests that emotions can systematically distort people’s receptiveness to advice and thus their rationality. And if everyone errs in similar ways, that could skew the classicists’ perfect calculus. “My intuition was that we often base complicated decisions on how we feel,” Schweitzer says. “If I ask you something complicated like, ‘Should we hire this person or should we buy this house?’ you have to consider a lot of attributes and compare a lot of complex things. So we often use a simple summary statistic, which is how we feel about the job candidate or the house. When we do that, we open ourselves up to the possibility of making a mistake based on emotion.”

That makes sense, but how do you prove it? Schweitzer and Gino designed experiments in which they — as difficult as it sounds — manipulated their subjects’ emotions, gave them advice and measured the effects. In their first experiment, they recruited college students and asked them to make a judgment about something they were sure they could not know for certain. In this case, they showed each subject a photograph of another person and asked them to estimate the body weight of the person in the photo. They then induced an emotion by having each subject watch a short movie clip. Some subjects saw an anger-inducing bit from The Bodyguard in which a man gets treated unfairly. Others viewed a gratitude-inducing clip from Awakenings in which another man receives an unexpected favor from his co-workers. And the rest saw a neutral outtake from a National Geographic documentary about Australia’s Great Barrier Reef.

In a separate study, the two scholars assessed how the videos induced different emotions. Because the students had no real connection to the scenes, the researchers could classify their reactions as incidental as opposed to integral. If you watch The Sopranos and then get angry with your spouse, that’s incidental emotion. If your spouse slaps you and you get angry with your spouse, that’s integral.

After watching the clips, the students reflected in writing on what they had seen and how it had made them feel, and then had a chance to re-estimate the weights of the people in the pictures. This time, they also received estimates that the researchers told them had been done by another participant. Though the subjects didn’t know it, everyone received the same set of second estimates. These estimates — the advice — were helpful, not misleading. “The emotion manipulations significantly influenced the accuracy of participants’ final estimates,” the two scholars state.

Participants “who experienced incidental gratitude weighed advice more heavily than did participants in a neutral state,” they write. “Participants who experienced incidental anger weighed advice less heavily than did participants in a neutral state. Even though the emotions induced in this study were unrelated to the judgment task, we find that these emotions significantly changed the extent to which participants relied upon advice.”

In the real world, as opposed to a behavioral lab, these findings play out in all sorts of ways. Co-workers, for example, often annoy each other, sometimes for legitimate reasons, like missed deadlines, and sometimes for silly ones, like how stupid someone’s laugh sounds. And sometimes, a person will get ticked off and fail to heed another’s good counsel just because of a bad mood.

“If I’m angry at my wife and therefore trust you less and am less receptive to your advice, then that’s clearly irrational,” Schweitzer says. “The fact that my wife crashed my car has nothing to do with you. But maybe I’m angry because you cancelled our last meeting and now we’re interacting again. Maybe there’s some real information about your reliability in the fact that you cancelled our meeting. It takes a controlled, clean experiment to disentangle rational reasons from biased ones. What we haven’t shown [with this study] but I’m confident would work is that, if you do something that makes me angry, then I trust your advice differently.”

Schweitzer says that people with what he calls “high emotional intelligence” are probably already putting his and Gino’s insights into action without even knowing it. “Emotional intelligence is the ability to recognize emotions and understand how they operate and also the ability to manipulate or change them. If I have emotional intelligence, I know what the right time to talk to my boss is. I know that my new partners had a terrible flight and lost their luggage and aren’t going to be receptive to what I’m saying, so I shouldn’t make my pitch right now. Or I know that, if I take them to this particular restaurant or I buy tickets to this Indy car race, I can shift their emotional state to feeling more gratitude toward me and listening to me.”

Skilled negotiators tend to have high levels of this kind of aptitude, and they apply it in small, subtle ways when they are doing their work. They might, for example, apologize for a perceived wrong, even when no apology was expected or required. Or they might, during a particularly tense time, call for a break, go get a soda and also bring something back for the people on the other side of the table.

Schweitzer sees what he and Gino observed operating in all sorts of business interactions. When a sales person takes a client to a ball game, for example, he’s not just cozying up in the obvious way. He’s also creating a sense of gratitude. When a drug rep brings lunch to a doctor’s office, she’s doing the same thing. “Can this backfire?” he asks. “Yes. If it doesn’t seem genuine, people aren’t going to believe it. Suppose that I try to induce gratitude and I go over the top. That’s the sales rep who’s giving too many gifts.” Push it too far, in other words, and you could end up making someone angry.

Observations On Patient Compliance Articles Not Presented As Patient Compliance Articles

Those accustomed to reading about patient compliance in publications such as The New England Journal Of Medicine, The American Journal of Psychiatry, The American Journal of Managed Care, white papers put out by pharmaceutical manufacturers and benefits management companies, and, of course, AlignMap.com, may find my free form observations helpful in orienting themselves in this brave new world.

1. The referenced article does not mention healthcare but does list an extensive set of business scenarios in which emotional content could affect ones decisions. The application of the content to compliance seems, as I read it, strikingly apparent. This is not, in my experience, unusual. Literature with a business, sociology, or economics orientation, for example, seem less concerned about how decision-making (in this case) works in specific, well defined situations than finding general principles that are valid in many settings. When healthcare is mentioned, it is often as one of many examples.
2. The article’s primary finding, that emotions experienced by the individual affect how that person responds to advice, even if the origin of those emotions have nothing to do with the immediate decision to be made – or, to extrapolate, the patient’s disorder or the healthcare situation, has not been emphasized in the medical literature.² Although in this example the findings are only moderately different from the conclusions of analogous articles with medical orientations, other instances will demonstrate entirely different, but not necessarily contradictory, approaches.
3. The experiments designed to test the hypothesis in this article lie closer to the basic research pole of the pure science-applied science spectrum than do the typical patient compliance studies and, not incidentally, are more akin to the animal behavior labs than naturalistic clinical trials favored in healthcare journals. Experimental approaches to similar questions vary dramatically from field to field.

My contention is not that the compliance-pertinent material available from non-medical fields is of higher (or lower) quality, that its experimental style is more (or less) valid, or that its findings are more (or less) useful. My contention is that the work done in non-medical fields often asks different questions, approaches solutions differently, presents findings in different contexts, … .³

Given the lack of progress in comprehending the workings of, let alone improving, compliance after many years of effort by the mainstream healthcare fields, the exploration of the potential contributions from these legitimate, well credentialed alternatives would seem a wise investment, if not an obligation, for anyone invested in understanding the phenomenon that most of us know by its healthcare-names, patient compliance or treatment adherence.


Footnotes

1. Re the more optimistic tone of today’s post, not to worry; this blog’s normal apocalyptic programming will resume forthwith↩
2. There has been significant material published in the medical compliance literature on stress caused by the medical problem being treated, the coping styles of the patients, and co-existing psychiatric diagnoses, especially depression.↩
3. Research and theoretical work in each of these non-healthcare fields may be as restrictive and narrowly focused as that done in healthcare. I am only pointing out that these fields view and treat issues that are part and parcel of patient compliance differently than do those of us in medicine.↩

Source: Great(er) Expectations John Lauriello, M.D., Am J Psychiatry, 164:377-379, March 2007

Expectations For This Post

The following excerpts are from an editorial written by John Lauriello, M.D. in the March 2007 American Journal of Psychiatry. While I have selected those portions that are most pertinent to the goals of treatment for schizophrenia and patient compliance, the entire essay is accessible, well written, and insightful – it is, all in all, a worthy read.

I have little to add to Dr Lauriello’s thoughts on these issues. Consequently, I offer these excerpts with no further elaboration.

At least for now.

“Take nothing on its looks; take everything on evidence. There’s no better rule.”

—Charles Dickens, Great Expectations

Although the successful treatment of schizophrenia is most often measured by symptom reduction and relapse prevention, the quality of everyday life and the ability to function independently are equally important to patients and families. To date, antipsychotic medications have not yet been shown to directly impact quality of life, particularly social and vocational functioning, to the degree hoped. Clinicians are challenged to prescribe medications that not only reduce symptoms but somehow also enable patients to become better functioning members of society. Presumably, as positive and negative symptoms of schizophrenia subside, subjective quality of life and objective measures of psychosocial functioning might be expected to improve. In this issue, two interesting and somewhat complementary articles focus on quality of life and competitive work performance of patients with schizophrenia. Both examine the promises and the limits of our current means of improving psychosocial functioning in these patients.

In the first article, Marvin Swartz and colleagues expand on findings of NIMH’s Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study, presenting results derived from Quality of Life Scale ratings. The analysis concentrated on those patients who continued their assigned medications for at least 12 months (a third of the overall sample). …

Contrary to predictions, researchers found no evidence to favor any of the second-generation antipsychotic medications with regard to improving Quality of Life Scale scores. As expected, patients’ baseline quality of life ratings showed impairment, especially in the vocational domain. After 12 months, only modest improvements were associated with any of the randomly assigned medications, with no differences found among any of the individual medications. ……

The Swartz et al. study confirms that improving quality of life does not come prepackaged in a medication bottle. As these authors state, “More intensive psychosocial rehabilitative services, including cognitive rehabilitation, may be needed to affect more substantial gains in functioning…. For patients unable to work, with limited access to vocational and rehabilitative services, even optimal medication may not be reasonably expected to improve … community functioning.”

This statement provides a perfect segue to the second paper, in which Susan McGurk and colleagues examine the effect of cognitive training on supported employment services at two community health centers … . In this study a small number of patients with schizophrenia were randomly assigned to receive supported employment alone or supported employment with cognitive training with the goal of increasing competitive work. Competitive work, the gold standard of vocational functioning, means competing in the marketplace for real work and real wages. Rates of competitive employment for patients with schizophrenia range from 10%–20%. The cognitive training administered by trained cognitive specialists, the “Thinking Skills for Work Program,” consisted of an extensive analysis of the cognitive work difficulties, computer-based training, planning sessions, and ongoing on-the-job assistance. Both study sites were also rated on the fidelity of their adherence to the supported employment model utilizing a standardized scale.

… The good news was that those in the combined program (supported employment with cognitive training) were more likely to work, held more jobs, worked for a longer time, and made more money. An additional benefit to those randomly assigned to the combined strategy was an improvement in depressive and autistic preoccupations as measured by the Positive and Negative Syndrome Scale (PANSS) at 3 months. In other words, those who worked more were less depressed and less internally isolated. The not-so-good news was that despite these noteworthy efforts, the mean number of weeks worked over the 2-year follow-up was only 27 weeks, roughly 6 months. Working for a quarter of the evaluation period (significantly better than the 5 weeks worked by those in the supported employment alone group) is laudable but still falls far short of independent self-sufficiency. One further note: although both study sites embraced the concepts of supported employment and cognitive training, one site scored lower on “fidelity” to the supported employment model. In turn, patients at that site performed worse than those at the more adherent site.

So what do these two studies together tell us about the psychosocial functioning of patients with schizophrenia? The Swartz et al. study coupled with the results of other CATIE publications demonstrate that medication adherence is challenging and that no one medication seems to be a clear winner when balancing all factors. But medication adherence is the stable platform that reduces exacerbations and rehospitalization. And this stability is necessary before any sustained psychosocial treatment can be applied. Supported employment with cognitive training appears to be one good next step for stable patients. The McGurk et al. study shows that adding cognitive training significantly improved time at work and reduced depressive preoccupations and isolating thinking. However, as with medication treatment, functional gains were limited and clearly dependent on adherence to the treatment. (emphasis added)

Follow-up To AlignMap Post On Study Claiming Switch From Lipitor To Generic Is Dangerous

Rather than sacrifice blogging’s inherent advantage of timeliness, I’m deferring the promised conclusion and summary post on the Personal Medication Record to link to and comment on this 3 November 2007 New York Times article, Maker of Lipitor Digs In to Fight Generic Rival, by Stephanie Saul And Alex Berenson, which addresses the Pfizer-sponsored statin research project study that was the focus of the 8 September 2007 AlignMap post, Lipitor Would Rather Fight Than (Have Patients) Switch

The article from the Times recapitulates much of argument I made in the earlier blog entry sans my lengthy exposition of some of the study’s problematic methodological issues. The crux of the New York Times essay is congruent with the final line of my post:

The most useful lesson from this study may well be the reminder that certain medical research projects may have as much to do with market share as with morbidity and mortality.

I’ve included a short excerpt of the opening of the article but the full piece is available without charge at ~ Maker of Lipitor Digs In to Fight Generic Rival ~

Patient choice promotes adherence in preventive treatment for latent tuberculosis
T. W. Rennie, G. H. Bothamley, D. Engova, and I. P. Bates. Eur Respir J 2007; 30:728-735

The Study

This study’s abstract follows:

Commentary

Offering patients treatment choices may indeed promote adherence, but I’m not convinced the results of this study are evidence of this hypothesis.

First, the study is complex. The population included patients treated by one of two methods, 3 months of rifampicin and isoniazid or 6 months of isoniazid, over a six year period. During the first two years, the treatment for a given patient was chosen by the doctor (the reasons one method or the other was chosen were not provided) while the patients were offered their choice of treatments over the last four years of the study.

That the majority of patients (78%) chose the shorter treatment is hardly surprising. (Nor is it surprising that more patients completed a three month course of treatment than a six month course of treatment.) Since the split between 3 month treatments and 6 month treatments among all patients was almost equal (3 month treatment: 53% of patients; 6 month treatment: 47% if all patients) the majority of the physician-assigned patients must have received the longer treatment.

I suggest that conclusions based on comparisons between these inherently dissimilar groups are, at best, tenuous.

Further, as is often the case in compliance studies, the definitions of successful completion, potentially successful completion, and failure to complete are rational but somewhat arbitrary:

And, for the purposes of this study’s statistics,

While these decisions can be justified and, indeed, may be necessary, it is unlikely that the definitions of adherence and nonadherence used in this research are a perfect match for the notions of adherence and nonadherence used in everyday clinical practice.

There are other, less overt, problems with this study enroute to the conclusion that patient choice enhances compliance but these should be sufficient to cast doubt on the unequivocal declaration of the title. A shame, that.

The Story & The Study

Reading the 6 September 2007 Medication Noncompliance blog entry, Pfizer Funded Study Says: Don’t Stop Taking Your Lipitor!,¹ alerted me to an intriguing Pfizer-sponsored statin research project and the presentation of its findings carried in multiple print and internet outlets.

Points of Intrigue

What is there, one might ask, about this study of statins that could be called intriguing?

Well, one intriguing point is that the first dozen or so articles on a Google search for “Pfizer,” “Lipitor,” and “simvastatin” referencing this study contain almost identical text taken from a press release put out by Pfizer – which may explain why Reuters, the first source I found after reading the Medication Noncompliance post, published the article under the title, Patients Who Switched from Established Lipitor Therapy to Simvastatin Experienced a Significant 30 Percent Increase in Relative Risk of Cardiovascular Events or Death, New Observational Study Shows, a rather longish headline that nonetheless effectively conveys the pro-Pfizer interpretation of the results conveniently uncluttered by any messy details and disclaimers.

A second intriguing point is that, although the study would certainly be categorized as medical research, was presented at the European Society of Cardiology Congress, and is to be published in The British Journal of Cardiology, most of these outlets carrying news of the study positioned it in the investment and business divisions of their web sites, as indicated by the domain and subdomain names of the pertinent URLs assigned to the articles, including, among others, biz.yahoo.com, money.cnn.com, www.forbes.com, finance.google.com, today.reuters.com/stocks, news.moneycentral.msn.com/ticker, www.quote.com. The story also received considerable play in the newsletters and sites covering the pharmaceutical industry (e.g., Drug Industry Daily and Pharmaceutical Business Review). Reportage of the study at clinically oriented sites (e.g., WebMD – which did not quote extensively from the press release) was notably less extensive.

Clearly, there is nothing illegitimate about a consideration of the business implications of medical research findings (or, for that matter, distributing press releases on the occasion of favorable news for the sending entity); I only suggest that garnering an understanding of this study and others like it might be abetted if its commercial aspect and its marketing spin are taken into account.

The clinical and marketing motivations for the study almost assuredly originate from the ramifications of the same historical event: Zocor, the branded version of simvastatin, a statin found to have been an effective agent albeit one somewhat less powerful than Lipitor, lost its patent protection in June 2006. Consequently, the price of this medication dropped significantly, causing a significant fraction of Lipitor users to change their medication to generic simvastatin, a shift encouraged and called for by certain third party payers, physicians, and patients themselves.

Pfizer, unsurprisingly, has questioned the clinical wisdom of patients migrating from their product, Lipitor, to a generic drug.² For a congruent but more fiscally eloquent explanation, consider this excerpt from Fighting the Generics Next Door published 7 September 2007 at The Motley Fool:

“Observational study” means the researchers reviewed records of patients who in the past took Lipitor, some of whom switched to simvastatin. Retrospective studies of this sort typically lack certain desirable qualities of prospective studies, such as randomization of patients, control of variables, and complete pertinent information. In this study, as the report notes, “Reasons for why treatment was discontinued were not available from the database, Pfizer indicated, adding that reasons for switching from one drug to the other were also not available. Furthermore, Pfizer explained that patients in the study were not randomised to each arm, which limits the significance of the findings.”

The non-randomized sample means that the two comparison groups, Lipitor-Continuers and Lipitor-Switchers, were not necessarily comparable. One detail , for example, from the poster session that didn’t make its way into the press release is that, as reported in Cholesterol Drug Change May Be Risky, “as a group, the patients who remained on Lipitor also took more heart medications and had better cholesterol control to begin with than the patients who switched to simvastatin,” a fact that could go far to explain the difference in outcomes.

The Patient Compliance Angle

The significance of the conclusions was especially limited for me since my personal interest in this study arose from a secondary finding:

Without knowing why the treatments were discontinued or why the patients were switched, this finding is interesting in a “how about that?” sort of way, but almost worthless for generating hypotheses or useful clinical practices.

Possible reasons for the higher noncompliance rates among the patients who switched medications are easy and fun to generate; how about these three for starters:

1. Patients switched because of the cost of medication. Those same patients who switched from Lipitor might well have been those individuals most likely to discontinue medication altogether for financial reasons – even though if the cost of simvastatin were relatively less than that of Lipitor.³
2. Patients switched because of side-effects of Lipitor. Those same patients might well have proved more vulnerable or less able to tolerate side-effects of other statins as well, including simvastatin.
3. Patients switched because of unresponsiveness to Lipitor. The departure of these patients from the Lipitor-taking group would immediately improve the outcome statistics of that category of patients by eliminating their negative results. Moreover, they could well have been more likely to be unresponsive to any statin and thus increase the apparent cardiovascular risk of the “Switched To Simvastatin” group.

Also, while the number of patients who discontinued treatment altogether is given, the actual compliance rates of those who continued treatment with either drug isn’t provided. (In other words, it isn’t known how many patients who continued to be enrolled in treatment with Lipitor, for example, actually took clinically effective doses of that medication, how many took the medication erratically, and how many professed to be taking the medicine but, in fact, were taking none.) Consequently, the effectiveness or ineffectiveness of taking adequate doses of Lipitor and simvastatin is impossible to calculate with confidence.

The Lipitor Challenge

If the study’s design is a researcher’s nightmare, it is a marketer’s delight.

Remember the “Pepsi Challenge” ads that showed blind tastings of Pepsi and Coke by cola drinkers with a stated preference for Coca-Cola? The design of that survey virtually guaranteed a good outcome for Pepsi – if even 10% of those tested, a number probably within the statistical error of the study, had chosen Pepsi, it would have been a triumph for that brand since those were ostensibly hard-core Coke drinkers.

The design of the “Stayed With Lipitor Vs Switched To Simvastatin” study similarly stacks the odds in favor of Lipitor. For one thing, it’s generally acknowledged that Lipitor is more effective than simvastatin; the question is whether the difference justifies the higher cost. So, from the outset, it was clear patients on Lipitor should have had better outcomes. It also seems likely that those patients who remained on Lipitor rather than switching were responding adequately or at least their doctors saw no benefit in moving them to a different drug. As suggested before, patients who switched from Lipitor might well be more likely noncompliant with, vulnerable to side-effects of, or unresponsive to other statins.

But the real clinker to those who don’t follow statin clinical research may be the consensus of earlier studies that discontinuation of any statin increases cardiovascular risk.

For example, New evidence links discontinuation of statin therapy to increased death rates in stroke survivors reports on a study in Stroke: Journal of the American Heart Association that demonstrated that “patients who stop taking cholesterol-lowering drugs within a year of surviving a stroke had a two-fold increased risk of death.”

And Statin Withdrawal Hard on the Heart – Discontinuing Statins Can Lead to Rapid Rise in Cholesterol and C-Reactive Protein notes

So, there would seem to be a reasonable chance that if Merck & Co had sponsored research looking at patients who switched from Zocor (assuming Zocor were still under patent – and incredibly profitable) to a generic form of Lipitor (or any other statin), the headline could have been “Patients Who Switched from Established Zocor Therapy to Atorvastatin Experienced a Significant X Percent Increase in Relative Risk of Cardiovascular Events or Death, New Observational Study Shows.”

Now, that would be an intriguing study, eh?

Conclusions

Maybe there is something especially catastrophic about shifting from Lipitor to simvastatin – or maybe not. In any case, nothing in this research proves either side of the debate. The findings do seem in line with the concept that discontinuing statins places patients at much higher cardiovascular risk, which is helpful but hardly groundbreaking.

The most useful lesson from this study may well be the reminder that certain medical research projects may have as much to do with market share as with morbidity and mortality.

Footnotes

1. I view the take on this study espoused in the Medication Noncompliance post reasonable, informative, and, as always, interesting, but, on the whole, lacking in the kind of embittered cynicism characteristic of the AlignMap perspective. Welcome to the Dark Side.↩
2. I am not the originator of this concept; The New York Times, for example, in its October 15, 2005 article, Lipitor or Generic? Billion-Dollar Battle Looms, predicted this sort of thing almost a year before Zocor’s patent protection expired.↩
3. While relative cost of these medications is now a primary issue, this is an unlikely cause of switching in this study, given that the observational period was from October 1997 to June 2005, a period during which Zocor was still under patent and, presumably, did not cost markedly less than Lipitor.↩

Adherence to Treatment in Poorer Countries: A New Research Direction? Alison Breen, M.A., Leslie Swartz, Ph.D., John Joska, M.B.Ch.B., F.C.Psych.(S.A.), Alan J. Flisher, Ph.D., F.C.Psych.(S.A.) and Joanne Corrigall, M.B.B.Ch., D.M.H.(S.A.) Psychiatr Serv 58:567-568, April 2007

The Letter

This letter to the editor points out the obvious but often overlooked fact that “most studies of treatment adherence have been conducted in high-income countries and the question arises as to whether there are structural barriers to adherence that are particular to, but underresearched in, areas with poor resources.”

The authors looked at a qualitative study of ten cases conducted in South African households with members who were caring for a household member who had schizophrenia and living in poor urban environments. Several themes were identified indicative of the particular healthcare compliance problems that challenge the poor.

Commentary

The factors affecting compliance seem never-ending. Nonetheless, as the authors of this letter point out, poverty inflicts special problems on a large number of patients and research efforts in this area are essential – for all of us.

Progress Toward Diabetic Treatment Goals

Source: Diabetes Patients Make Significant Gains in Disease Control; Still, Nearly Half Fail to Reach ADA Treatment Goals

According to the Quest Diagnostics Health Trends Report on Diabetes,¹
presented on 23 June 2007 at the ADA Annual Scientific Conference (posters # 21-LB and 22-LB), Americans being treated for diabetes improved control of their disease over the last six years, by 44.4%.

Seasonality

This excerpt from the article address seasonality of compliance:

Commentary

While I am uncomfortable presenting an issue based exclusively on a press release from a company intimately involved in accumulating the data, I am struck by the seasonal variation in results, a trend rarely considered as a factor in patient compliance.

This excerpt touches on the implications of the findings:

Life course and seasonality may be as important in the context of adherence as they are in chronic disease. Certainly, this could be a rewarding focus for future research.

Footnotes

1. The Quest Diagnostics Health Trends Report on Diabetes is based on findings from 22.7 million de-identified HbA1c tests, a key indicator of diabetes control, performed by Quest Diagnostics between 2001 and 2006 on 4.8 million patients who saw a healthcare professional and were classified as having diabetes.↩