Chief Editor

Dr Abdulrazak Abyad
Abyad Medical Center & Middle East Longevity Institute
Azmi Street, Abdo Center,
PO BOX 618
Tripoli
Lebanon

Phone:
(961) 6-443684
Fax:
(961) 6-443685
Email: aabyad@cyberia.net.lb

Editorial Board

Past issues

Current Issues

Submission of articles

Advertising detail

Publishers
medi+WORLD International
572 Burwood Road
Hawthorn
Australia

Ph: +61 3 9819 1224
Fax: +61 3 9819 3269

Marcel GM Olde Rikkert
Wim JM Dekkers

Marcel GM Olde Rikkert, PhD MD, Geriatrician, Dept Geriatrics, University Medical Centre Nijmegen, The Netherlands

Wim JM Dekkers, MD PhD, physician and philosopher, Dept of Ethics, Philosophy and History of Medicine, University Medical Centre Nijmegen, The Netherlands

Abstract
In geriatrics there often exist dilemma's whether or not to apply treatments with the maximum theoretical benefit, as known from randomized clinical trials. One reason of doubt is that the trials are often carried out in younger patients. The choices in such dilemma's should not only be determined by medical facts of the treatment options, availability of alternative treatments, and extrapolation of chances for beneficial outcome and adverse reactions to geriatric patients, but also by experiences of the physician in charge both with the patient's medical problem and the treatment options, and last but not least the emotions of the patients concerning the options. Using a case history we state that the calculation of the individualized Likelihood of being Helped versus being Harmed (LHH)) does help to clarify all ingredients of such dilemma's. Pros, cons and limitations of LHH-estimation in geriatrics are discussed, as well as the possibility to quantify evidence, emotion and experience in the weighing up and individualizing of treatment choices for frail elderly patients.

Key words: Likelihood of being Helped versus being Harmed, ethics, geriatrics, informed consent, anticoagulants

Key messages

• Patients' emotions and physicians' experience in treatment dilemma's should be taken into account in Evidence Based Medicine.
• The Likelihood of being Helped versus being Harmed is a useful tool to translate general evidence in individualized decision making.
• Memory impairments do not preclude decision making capacity.

Consent
The family of mr. W. approved of the publication of this case history.

Introduction

Seventy-five years ago Francis W. Peabody wrote that the physician who attempts to care for a patient while neglecting his emotional life, is as unscientific as the investigator who neglects to control all the conditions that may affect his experiment.^[1] Protagonists of evidence based medicine (EBM) have adopted this wisdom into the calculation of the likelihood of being helped versus being harmed (LHH), which may guide us in therapeutic dilemma's.^[2] Physicians' clinical experience and their knowledge of risk factors in individual patients, can be taken into account in this LHH, by adapting the absolute risks for preventable outcome of disease or adverse events. Moreover, patients' emotions and experiences, expressed in the explicit weighing of positive outcome versus adverse reactions, are an important ingredient of LHH. We present a case history that illustrates that the LHH estimation can also be used in treatment dilemma's in frail elderly subjects, and that minor cognitive impairments without major problems in judgement do not preclude this. Evidence based geriatrics taking the LHH into account does not resolve treatment dilemma's, but may clarify them rationally.

Case history

Wehad known mr. W., born in 1916, already for more than ten years when we finally came entangled in a treatment dilemma, because he took part in several studies in our institute. He was exceptionally fit during his seventies and early eighties. Therefore, we were surprised to see him as an outpatient, because of memory complaints. After assessing him with our routine mental and physical investigations, we diagnosed a mild cognitive impairment. Among other tests, he had a score of 28 out of 30 on the MiniMental State Examination. Brain imaging by CT scanning showed no infarctions and only minimal peripheral atrophy. However, he did turn out to have atrial fibrillation, without a specific cause, which had been unknown so far. He and his wife felt greatly relieved after my reassurance that he did not suffer from dementia. We agreed that he would start with anticoagulant therapy, after we had informed him that he would have a nearly 70% risk reduction on a first stroke by taking coumarines, versus about half of this relative risk reduction on aspirin, at the cost of only a moderately increased risk for minor bleedings.^[3,4] We also told him that his overall risk on a stroke was low, because of the absence of cardiovascular morbidity.

Unfortunately, after nearly two years he did get a serious nose bleeding, for which he had to be admitted to hospital. Few days after admission, we saw him at the Ear-Nose-Throat department, because his bleeding had been complicated by a delirium. We did not find other causes for his delirium, and gave him a low dose of haloperidol for his hallucinations. After recovery from delirium he was discharged to his house, but after discharge he developed a depression, for which he needed antidepressants. During this cascade of events it became clear that he had become fragile, with an increased risk for further impairments.
From that moment on we faced the dilemma of again starting anticoagulant therapy or aspirine. Because of increasing cognitive impairment the question of whether or not to assess capacity to consent also forced itself into this dilemma.

At that time we still estimated his absolute stroke risk as moderate or low, because he did not suffer from hypertension, and he did not have stroke before, nor silent infarctions on his CT scan. We told him that his yearly risk for stroke with anticoagulant therapy would probably be around 5 %, while his stroke risk on aspirin would be around 5.5 %.^[4] The one thing he certainly did not want again at that moment, was a second bleeding like the one that caused his admission. He had become very anxious after what turned out to be his first medical emergency. Therefore, the lower risk for bleeding on aspirin, was much more important for him than the minor difference in risk reduction for stroke. Well informed, he chose for aspirin.

This dilemma of aspirin or anticoagulants intensified itself when he suffered a minor stroke few months later, after we had put him on aspirin.

Treatment options

The first option was to start again on aspirin, with a suboptimal risk reduction of recurrent stroke. The second option was to restart with anticoagulant therapy, with a better risk reduction for stroke at the cost of a higher chance for recurrent bleeding (4.2 % per patient per year on aspirin vs 11.8 % per patient per year on anticoagulant therapy).^[5] We also considered other treatment strategies that might make prescription of anticoagulant therapy unnecessary. Rhythm regulation was not considered to offer real advantages, as anticoagulant therapy is still recommended, regardless of whether the patient's rhythm is returned to sinus or remains in atrial fibrillation.^[6,7] Dipyridamol was not discussed either, because of limited data and unclear effect on bleeding risks.

Arguments and reasons
Implementing the best data on risks and treatment effects, together with patients' preferences is worth the effort, especially in treatment dilemma's, because weighing can be done more rationally and transparently. Calculating an individualized LHH, as part of evidence based medicine, serves this goal. In this case, emotional valuation of bleeding largely determined the LHH estimation.
Some comments can be made on the LHH procedure that was followed. For easiness, we directly asked for weighing of stroke versus bleeding. This was closest linked to the clinical dilemma. Risks and benefits can also be weighed separately.^[8] Mr W. concluded he would prefer stroke ten times over another bleeding. The figure was included in the LHH-formula:

In which there are three components:
Literature based:
NNT is the number needed to treat to prevent recurrent stroke (12 on anticoagulants versus 53 on aspirin)^[5]
NNH is the number needed to treat to cause one bleeding known (11 on anticoagulants versus 91 on aspirin)^[5]
Physician based:
FS is the experience based correction for the likelihood of stroke in this patient (i.e. 1)
FB is the experience based correction for the likelihood of bleeding in this patient (i.e.: 3 for anticoagulant therapy; 1 for aspirin)
Patient based:
S is the preference of prevention of stroke versus prevention of bleeding (i.e.: 0.10)

The LHH-calculation was made two times, both for anticoagulant therapy and aspirin. One can even combine these two in one expression of the LHH for coumarines versus aspirin. However, the result may become too complex, thereby losing the transparency that was aimed for.

There are some limitations of the formal calculation of the LHH in frail elderly patients. In case of progressive cognitive impairment, patients will finally lose the ability to weigh alternatives accurately. One may argue that this is the reason why prescription of anti-dementia drugs for demented patients can never really be evidence based practice.^[9] However, in general it is always of advantage for the patient when his preferences of care are seriously investigated. Calculating the LHH stimulates this.

The other precondition of having reliable data on positive outcome and adverse events for geriatric patients, will often not be met. Recently, it was shown that from 1990-2002 only 84 randomised controlled trials (RCT) on patients of 80 years and over were published.^[10] One can estimate this as less than a half percent of all RCTs carried out in that period. This means that we often must extrapolate data for geriatric patients. It is unknown and probably depending on the issue studied, what error will be introduced in this way. However, for the comparison of treatment options in atrial fibrillation sufficient good quality data are available, which are partly collected in patients in their eighties. This patient was especially well served, because the absolute risk percentages and the data on secondary prevention therapy and bleedings risks came from a European study (EAFT). ^[5] This study even collected the data partly in The Netherlands. Mr. W. might have been included in the study, had his history played earlier. Data on the risk of rebleeding, however, were not available, and had to be estimated from data on first bleeding risks.

Practical conclusions

One week after his minor right hemispheral stroke, he was fully recovered of the paresis of his left hand and arm, and the subtle loss of muscle power in his left leg. His memory impairment increased. He had a score of 23 out of 30 on the MiniMental State Examination. Phasis, praxis, gnosis, insight, judgement and decision making capacity were not affected, so he did not fulfil the criteria for a dementia syndrome. His mood was not affected, though he realized he had definitely lost his former strength. Mr W. had now migrated to a higher risk category for having another stroke, which was estimated as 12% per year in case of neither aspirin nor anticoagulant therapy.^[5] The debate whether to start again with aspirin, or change again to anticoagulant therapy had become a real dilemma.

After careful discussion with mr W., we calculated the LHH for anticoagulant therapy and for aspirin, estimating the risk of a bleeding on anticoagulants three times higher for mr. W. than average, because he already had suffered a bleeding on anticoagulant therapy. His individualized LHH for coumarines was 1 against 32. For aspirin his overall LHH turned out to be 1 against 6. He had more chance on coumarines on a negative effect of his treatment. After our discussion, he did not change his preference for aspirin, though he admitted it for a major part this was based on his fear rebleeding, in which the complications of delirium and depression versus the reversibility of his stroke were recognized as important factors.

He did not change this opinion during the next days, in which we judged him as capable to decide for himself on this issue. After a week hospitalisation, he was discharged on aspirin, to which we also added perindopril and indapamin, as evidence based secondary prevention measures preventing stoke for patients over 70.^[11]

Only three months later mr. W. got another cerebral infarction, this time a major left hemispheral ischemic event. His consciousness declined, he got fever, and a hypomotoric delirium. Added to his right-sided hemiplegia, he could not swallow without aspiration. There was only little recovery in the first week. In agreement with his family, no tube feeding was started. Mr. W. died peacefully a short time afterwards, aged 88 years. His family was happy with the way he aged, died, and was cared for.

Conclusion

This case clearly unerlines that also in geriatric treatment dilemmas the calculation of LHH may offer the benefit of maximal clarity and explicit communication on emotional preferences. While in general elderly clearly benefit most from anticoagulant therapy in case of atrial fibrillation, we argue that in individual cases the patients' preferences may be different, but still have to guide us. Minor cognitive impairment is no reason for dismissing these emotion based opinions. If (pathological) fears or emotions are not so strong that a patient is considered incapable to consent, these elements should be weighed similarly to the rational benefits of treatments.
Referring to the case we feel that one cannot judge the quality of medical decision making on the final outcome only. This seems obvious, but has important consequences, for example in medical audits. We all experience patients, who, while knowing all the information, prefer an emotional rather than a rational decision.^[12] As a consequence, audit on the quality of prescription should also examine the medical record, rather than just the prescription card or the outcome of the intervention.

The core of geriatrics is to develop an individualized bio-psycho-social treatment plan, taking into account the fragility of the patient. Assessment of clear treatment goals related to the daily life and preferences of care of the patient is crucial for a successful intervention. Formal methods, such as the calculation of the LHH are also useful in treatment dilemma's in competent geriatric patients. LHH will help geriatricians, who often have to chose between what may seem apples and pears: evidence based pros and cons on the one hand, and patients' and physicians' emotions and experiences on the other.

References

November 2004
Volume 1,
Issue 2


Table of Contents

Home

Editorial Ageing

Meet the team

Blood RNA concentration in Alzheimer disease and vascular dementia

Serum Zinc and Copper Concentration in Human-Age related cataract

The likelihood of being helped versus being harmed: useful in geriatric treatment dilemma's

Feasibility of the Divided Attention Steering Simulator (DASS) for the Assessment of Vigilance in Stroke Patients

The European Nursing Academy for Care of Older persons (ENACO)

Middle-East Academy for Medicine of Ageing, third session of the first course

Urogenital atrophy in climacteric women: Menopause or Geripause?

The diagnosis and management of dementia

The length of hospital stay of Home Health Care patients at King Khalid National Guard Hospital, Jeddah, 1999.

Congratulation!