Effectiveness of telephone counselling by a pharmacist in reducing mortality in patients receiving polypharmacy: randomised controlled trial
BMJ 2006; 333 doi: https://doi.org/10.1136/bmj.38905.447118.2F (Published 07 September 2006) Cite this as: BMJ 2006;333:522All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
I would like to thank Wu and colleagues (1) for an interesting and,
on the whole, well designed, study, which offered some challenging results
and became the choice article for critical appraisal at our Public Health
Directorate's journal club. We, like other responders, were "impressed by
effort"(2), but have the following comments.
Firstly, the question the authors were attempting to answer lacks
total clarity. They look at the effect of telephone counselling and
initial compliance on mortality, but are the authors implying a direct
causal pathway of telephone counselling - increased compliance - decreased
mortality? Although there were hints of this throughout the article, it
was never clearly stated.
Secondly, if randomisation had been carried out at screening, the 60
defaulters, who appeared to be of a different group, would have been
allocated to groups and included in the intention-to-treat analysis
possibly resulting in a more robust trial.
Thirdly, how do the authors account for the significant differences
in compliance score (Table 1) and the proportion of patients using
antiplatelets agents and lipid lowering drugs (Table 2) between the groups
at the start of the trial after randomisation?
Fourthly, the validity and reliability of the method of assessing
compliance used in the trial seemed somewhat uncertain. Tablet counts and
diaries as methods of measuring compliance have been seen by some as being
"inadequate", often leading to an overestimation of compliance (3).
Studies have shown that microelectronic monitoring, for example, may be a
more effective means of measuring compliance (4).
I was very impressed that the initial 15 minute talk held by the
pharmacist at the screening stage resulted in 58% of the non-intervention
group becoming and remaining compliant at 2 years (Figure 1) and believe
that the main take home message to clinicians must surely be: "talk to the
patient".
References:
1.Wu JYF, Leung WYS, Chang S, Lee B, Tong PCY, Chan JC. Effectiveness
of telephone counselling by a pharmacist in reducing mortality in patients
receiving polypharmacy: randomised controlled trial. BMJ 2006:333;522-527
2.Sumit, R. Majumdar. Impressed by effort ¨C skeptical of results.
BMJ Rapid Responses 10 September 2006.
3.Pullar T, Kumar S, Tindall H, Feely M. Time to stop counting
tablets? Clin Pharmacol Ther 1989;46:163-8.
4.Cramer JA, Rosenheck R. Compliance with medication regimens for
mental and physical disorders. Psychiatr Serv. 1998 Feb;49(2):196-201.
I would like to thank colleagues in the former Suffolk West PCT for
their helpful comments.
Competing interests:
None declared
Competing interests: No competing interests
We have some questions related to the paper on the effects of
telephone counselling (Wu JY et al., BMJ. 2006 Sep 9;333:522).
(1) For calculating the effects of telephone counselling on patient
compliance the authors included persons who died in the analyses.
Prevalence rates and statistical analyses of differences of compliance
rates between intervention and control groups should be based on surviving
subjects only. What are the effects if analyses are based on surviving
participants?
(2) The proportion of compliant persons at the two-year follow-up is
reported for subgroups of participants only, and not for the entire
intervention versus entire control group. Also average compliance scores
of intervention and control groups at the two year follow-up, with
corresponding statistical analyses, are not reported. We would be
interested to know these results.
(3) Based on manual addition of the number of compliant persons at
the two-year follow-up, 179 of 194 surviving persons of the intervention
group and 134 of 185 surviving persons of the control group were
compliant. The fact that more than 70% of persons of the control group
were compliant at the two-year follow-up is very surprising. Was there a
contamination effect or a secular trend?
Competing interests:
None declared
Competing interests: No competing interests
The study by Wu et al showed that a telephone service by pharmacists in Hong Kong could improve adherence and reduce mortality. Readers in other countries may wonder whether this work is generalizable across different cultures and health care systems. We can provide some evidence that it is, and that the service is cost effective.
We have conducted a similar study in the UK 1 targeted at people starting chronic therapy, when non-adherence can develop rapidly2. It was a theory driven, randomized controlled trial of 500 patients, the intervention group receiving a telephone call from a pharmacist 2 weeks after starting a new medicine for a chronic condition. At 4-week follow-up non adherence was significantly lower in the intervention group compared to control (9% vs 16%, p=0.03), as were the number of patients reporting problems with their medicines (23% vs 34%, p=0.02). Patients’ attitudes towards their medicines (known to be related to adherence3,4) were measured by the ‘necessity-concerns differential’3 (an indicator of how the patient judges their perceived need for the medication relative to their concerns about potential adverse effects). The differential was significantly more positive in the intervention group when compared to control (5.0 vs 3.5, p=0.007). Phone calls took a median of 12 minutes.
An economic evaluation, led by Elliott, will be published separately; the cost effectiveness acceptability curve shows there is an 86% chance that the intervention is dominant - less costly and more effective than normal treatment. In short, we would expect the intervention to be cost effective in the UK.
References
1. Clifford S, Barber N, Elliott R, Hartley E, Horne R. Patient centred advice is effective in improving adherence to medicines. Pharmacy World and Science (2006); e-publication DOI 10.1007/s11096-006-9026-6 Available from: http://www.springerlink.com/content/m31612w6648l6722/ [accessed 28th September 2006]
2. Barber,N., Parsons,J., Clifford,S., Darracott,R., and Horne,R., Patients' problems with new medication for chronic conditions. Quality and Safety in Health Care (2004); 13: 172-5.
3. Horne R, Weinman J. Patients' beliefs about prescribed medicines and their role inadherence to treatment in chronic physical illness - processes and applications. Journal of Psychosomatic Research (1999); 47(6):555-567.
4. Horne R, Weinman J. Self regulation and self management in asthma: Exploring the role of illness perceptions and treatment beliefs in explaining non-adherence to preventer medication. Psychology and Health (2002); 17(1):17-32.
Competing interests:
None declared
Competing interests: No competing interests
Wu and colleagues are to be congratulated on rigorously studying an
important question related to the role of pharmacists in improving patient
-related outcomes. This is particularly important in my own jurisdiction
(Alberta, Canada), where pharmacists have been accorded prescribing
privileges and are aggressively lobbying for “cognitive” fees and
alternate payment plans. Thus, any randomized trial that shows a 41%
reduction in all cause mortality by such a simple pharmacist-based
intervention deserves very close scrutiny. While impressed by Wu et al’s
efforts, I am somewhat skeptical of the results, for at least the
following 3 reasons:
First, the authors do not present unadjusted results for their main
outcome, mortality. Why conduct a randomized trial if you are going to
adjust your findings? In my mind, this basically undoes the benefits of
randomization and converts the trial into a very rigorous observational
study. I could not calculate an unadjusted log-rank test p-value without
the actual data, but (acknowledging slight issues of statistical
efficiency with person-time information) when I constructed a simple 2by2
table with data presented by the authors, the primary result was a
relative risk reduction of 0.67, 95% CI 0.42 to 1.07, p=0.12. At least by
this simplistic but transparent approach, the study could have been
associated with no benefit or even harm.
Second, perhaps the authors had always intended (a priori) that the
primary results would be based on “adjusted” results. Other than for
Table 1/2 imbalances (we are told the only differences were in baseline
compliance and use of statins and antiplatelet drugs), there is little
post hoc justification for such an approach, and the authors have adjusted
for a whole host of (unjustified, in the sense that randomization should
have balanced these things and leveled the playing field) covariates
(their Table 4). Concerns about adjusted analyses as the primary analysis
for a randomized trial can be understood by simply examining the other
results from Table 4. Specifically, the use of statins was associated
with a 72% reduction in all-cause 2-year mortality (p=0.02) in this study,
a benefit at least 3 times as large as that seen in the randomized trials
of statins [1]. Fortunately, this trial was registered at ISRCTN
[#48076318], which would permit me to better understand the study in the
absence of formal design paper. Unfortunately, it was registered in 2004
(when less information was required) and of the 15 fields required by
ISRCTN at the time, 5 were listed as “not available” – including things
like exclusion criteria, sample size, and outcome. Apparently, it is not
only the pharmaceutical industry that poorly registers trials[2].
Third, the effect size is remarkable. Acknowledging the results are
based on a difference of 13 events over 2 years, one rarely encounters a
valid 41% reduction in mortality in the biomedical literature of complex
interventions. Most of the benefit in Wu et al’s study appears to be
related to cardiovascular deaths. But the mortality benefits of telephone
advice (mostly from nurses, to be fair, but often pharmacists as well) to
improve adherence and processes of care have previously been reported to
be less than half of this size for coronary heart disease [3] and small-to
-absent for heart failure [4]. Interestingly, the authors cite the design
paper for the latter reference as the type of study needed to confirm and
replicate their work, rather than the non-confirmatory and robustly
negative for mortality results of DIAL[4].
In summary, I would encourage the authors to provide the unadjusted
results of the study or even provide the data to interested parties to
assure themselves of the robustness of the main study findings and to
finish the trial registration (current standards) completely and
accurately. And I would urge providers and policy makers to take these
more-than-impressive results (exposing 16 patients to a pharmacist
telephone call resulting in 1 life saved) with a grain of salt. Results
that are too-good-to-be-true are often too good to be true[5].
Respectfully submitted,
Sumit R. Majumdar, MD MPH FRCPC
REFERENCES
1. Cholesterol Treatment Trialists’ Collaborators. Efficacy and
safety of cholesterol-lowering treatment: meta-analysis of data from
90,506 participants in 14 randomised trials of statins. Lancet. 2005;
366:1267-78
2. Drazen JM, Wood AJ. Trial registration report card. N Engl J
Med. 2005; 353:2809-11
3. McAlister FA, Lawson F, Teo KK, Armstrong PW. Randomised trials
of secondary prevention programmes in coronary heart disease: systematic
review. BMJ. 2001;323:957-962
4. GESICA Investigators. Randomized trial of telephone intervention
in chronic heart failure: DIAL trial. BMJ. 2005;331:420-427
5. Sylvestre MP, Huszti E, Hanley JA. Do Oscar winners live longer
than less successful peers? A reanalysis of the evidence. Ann Intern Med.
2006;145:361-363
Competing interests:
None declared
Competing interests: No competing interests
Dear Editor,
I enjoyed reading this article by Jennifer y.f.Yu Et al.It is a very
important message and it highlights the importance of patients education.
We as clincians sometims focus our energy and attention mainly on acute
problems but fail to realise that patient's understanding of disease and
treatment is very important for the outcome.
As sited in this study the importance of pharmacy advice on reducing
mortality is another example of this. In United kingdom we have a
community based health care system and I feel it would be great idea that
if G.P'S could arrange sessions for patients for advice on medications and
their sideeffects, importance of compliance and its implications etc.
I feel the learning message from this article is that regular motivation
for compliance and patient education is the key to improve outcomes for
patients on multiple medications.
Kind Regards
Girish Chawla
Competing interests:
None declared
Competing interests: No competing interests
error in abstract
Dear Editor,
in abstract (end of Results section) it is said that
"In the cohort of 1011 patients, the adjusted
relative risk for death was 1.61 (1.05 to 2.48; P = 0.029) and 2.87 (1.80
to 2.57; P < 0.001) in patients with compliance scores of 34-66% and 0-
33%, respectively, compared with those who had a compliance score of 67%
or more."
Of course for 2.87 DI is not from 1.80 to 2.57, but to 4.57, as it is said
in the text.
I believe that abstract need to be corrected as it is freely available on
the web.
Competing interests:
None declared
Competing interests: No competing interests