Drugstop.co.il

Original Articles The war against Polypharmacy: A New Cost-Effective
Geriatric-Palliative Approach for Improving Drug
Therapy in Disabled Elderly People
Doron Garfinkel MD1, Sarah Zur-Gil MA2 and Joshua Ben-Israel MD3 1Department of Evaluation & Rehabilitation, 2Pharmacy, and 3Directorate, Shoham Geriatric Medical Center, Pardes Hana, Israel Key words: polypharmacy, geriatric-palliative approach, nursing departments/nursing homes
for clinical and laboratory changes, with the aim of improving quality of care.
Background: The extent of medical and financial problems of
polypharmacy in the elderly is disturbing, particularly in nursing homes and nursing departments. Patients and Methods
Objectives: To improve drug therapy and minimize drug intake
The study was conducted at the Shoham Geriatric Medical Center in nursing departments. in Israel. In early 2004, all patients in six nursing departments Methods: We introduced a geriatric-palliative approach and
(study departments) were evaluated by one of the authors methodology to combat the problem of polypharmacy. The study (D.G.) for all drugs consumed. An attempt was made to stop group comprised 119 disabled patients in six geriatric nursing departments; the control group included 71 patients of comparable as many drugs as possible, using the criteria of our geriatric- age, gender and co-morbidities in the same wards. After 12 months, palliative methodology [Figure 1]. The control group comprised we assessed whether any change in medications affected the death patients hospitalized in the same departments and treated by rate, referrals to acute care facility, and costs.
the same team, in whom no change in drugs was made. The Results: A total of 332 different drugs were discontinued
department physicians had complete authority to re-administer in 119 patients (average of 2.8 drugs per patient) and was not associated with significant adverse effects. The overall rate of drugs whenever drug discontinuation was defined as "failure" (see drug discontinuation failure was 18% of all patients and 10% of below). The algorithm in Figure 1 summarizes our methodology all drugs. The 1 year mortality rate was 45% in the control group for implementing the geriatric-palliative approach in nursing but only 21% in the study group (P < 0.001, chi-square test). The homes and nursing departments. It was used to reevaluate each patients' annual referral rate to acute care facilities was 30% in medication for each patient, enabling us to decide whether to the control group but only 11.8% in the study group (P < 0.002). The intervention was associated with a substantial decrease in continue with the same dose, reduce it, or discontinue the drug the cost of drugs. completely. When no evidence-based data were available for Conclusions: Application of the geriatric-palliative methodology
answering the first statement, we based our answers solely on in the disabled elderly enables simultaneous discontinuation of clinical judgment. If the indication seemed relevant in disabled several medications and yields a number of benefits: reduction in elders, we would have nevertheless considered dose reduction or mortality rates and referrals to acute care facilities, lower costs, and improved quality of living.
shift to a better drug while carefully monitoring for any change in symptoms, signs or relevant tests. Discontinuation of nitrates was tried in patients who had no chest pain for 3 months; failure was defined as the return The rate of drug-related problems and inappropriate medication of symptoms or electrocardiographic changes. H2 blockers were use in the elderly is disturbing. The heavy use of medications stopped in patients with no proven peptic ulcer, gastrointestinal in this population has increased the rate of drug interactions bleeding or dyspepsia for 1 year; failure was defined as the and hospitalizations secondary to drug-related problems [1,2]. return of upper gastrointestinal symptoms. Discontinuation of The extent of the problem is even greater in nursing home and potassium and iron supplements was tried in patients with serum nursing department settings [3,4], and the financial consequences concentrations above 4.0 mEq/L or 80 μg/dl, respectively. Failure of the problem are enormous.
was defined as a reduction in serum potassium below 3.5 mEq/L We introduced a geriatric-palliative approach and methodol- and that of iron below 50 μg/dl. When several antihypertensive ogy to improve the quality of care in nursing home/nursing drugs were consumed, we would try to remove only one while departments, assuming a priori that each patient in our maintaining the dosage of other antihypertensive drugs. Failure nursing department suffered from some negative effects of was defined as an increase in diastolic blood pressure above 90 polypharmacy. Our research hypothesis was that, in most pa- mmHg and/or systolic blood pressure above 140 mmHg. If suc- tients, several drugs could be discontinued without significant cessful, other antihypertensive drugs would be stopped according negative effects on mortality, morbidity and quality of life, and to the same principles. Whenever a specific drug discontinua- with beneficial financial consequences. In the present study tion was defined as "failure," the drug was re-administered. The we discontinued as many drugs as possible while monitoring success rate was determined 12 months after the intervention. D. Garfinkel et al.
• Vol 9 • June 2007
Original Articles Table 1. Demography and co-morbidities
An evidence-based consensus exists for using the drug for the indication given in its current dosing rate, Study group
Control group
in this patient's age group and disability level, and the benefit outweigh all possible known adverse effects Age (yrs) (mean ± SD) * Indication seems valid and relevant Double incontinence in this patient's age group and disability level Indwelling urinary catheter Congestive heart failure Do the known possible adverse reactions Previous myocardial infarction of the drug outweigh possible benefit Chronic atrial fibrillation in old, disabled patients? Diabetes mellitus Chronic obstructive lung disease Any adverse symptoms or signs Hypo-albuminemia (serum albumin < that may be related to the drug? Recurrent infections *** Another drug that may be superior All parameters except age, in both the study and control groups, were analyzed by the to the one in question chi-square test.
* Student's t-test Mini Mental State Examination (MMSE) 14/30 or less.
*** At least two proven infections in one year (urinary tract infection, pneumonia, skin Can the dosing rate be reduced with no significant risk? Table 2. Success rate following 1 year of follow-up
according to number of drugs discontinued CONTINUE WITH THE SAME DOSING RATE Failure rate:
Figure 1. Improving drug therapy in disabled/frail elderly patients
No. of drugs
At that time, the annual incidence of deaths and referrals to hospitals was determined in both the study and control groups.
All data were analyzed by the chi-square test. The average age was analyzed using Student's t-test. Unfortunately, we could not reliably compare the cost of drugs for patients in whom drug discontinuation was and was not performed in the same six nurs- ing departments. We therefore compared the cost of drugs in the six study departments (both study and control groups) to that of another four nursing departments in the same medical center (control departments), between January and July one year earlier, and the same period after the intervention (chi-square test).
types. The discontinuation of nitrates in 22 patients was not associated with any clinical or ECG changes; discontinuing H blockers did not cause upper gastrointestinal symptoms in 94% We evaluated the use of medications in 190 patients in the of patients; and discontinuation of antihypertensive drugs did not six study nursing departments. Drugs were discontinued in 119 cause an increase in blood pressure in 42 of 51 patients (82%). (63%); there was no change of medications in 71. The groups Furthermore, in nine patients defined as "failures," the number of were comparable for age, gender and major co-morbidities [Table antihypertensive medications or their dosage was reduced. The 1]. The average number of medications consumed was 7.09. success rate for pentoxyfyllin, potassium and iron supplements Altogether, 332 different drugs were discontinued (an average of was also remarkable. The failure rate of the geriatric-palliative 2.8 drugs per patient). The rate of successful drug discontinuation approach was highest for antidepressants and psychotropic drugs decreased as the number of discontinued drugs in one patient [Table 3]. Other drugs were discontinued (e.g., non-steroidal increased; the overall failure rate was 18% of all patients and anti-inflammatories, analgesics, statins, oral hypoglycemics, 10% of all drugs [Table 2]. amantadine, carbamazepine and digoxin), with no adverse find- Table 3 presents the annual rate of success by different drug ings that could be attributed to drug discontinuation. Due to the • Vol 9 • June 2007
Resolving Polypharmacy in the Disabled Elderly Original Articles Table 3. Success rate after 1 year of fol ow-up according to types of
Residents in nursing homes or nursing departments use an drugs discontinued average of 6 to 9.7 medications daily (7.09 in our study) and Recurrence of
over 20% receive more than 10 medications daily [9,10]. The rate No. of patients with
of drug-related problems in these settings is significantly higher Drug group
(failures)
than in community-dwelling elders [4,11,12]. Polypharmacy is preferably defined as "The administration of more medications than are clinically indicated" [13]. Another term is "inappropriate medication use" – medication use that has a greater potential Diuretics (furosemide) risk for harm than benefit, is less effective or more costly than available alternatives, or does not agree with accepted medical Potassium supplement standards. However, there is still considerable disagreement among experts regarding what exactly is inappropriate medication Sedatives & tranquilizers use and how it can be determined [14].
Beers et al. [10,15] tried to establish criteria for defining groups of drugs or specific medications that should be regarded * See text for further explanations as "potentially inappropriate" and should not be given to elders in nursing homes or nursing departments. Chutka and colleagues [16] claimed that there was insufficient evidence to conclusively small number of patients, statistical analysis was not performed defend or refute the use of most medications listed by Beers. for these drugs. In some patients in the study group, the staff This uncertainty may explain the different incidence of inappro- reported decreased agitation, increased alertness and even an priate medication use reported by many researchers in the com- amelioration of disability, but we did not quantitatively assess munity [1,3,14,17-19] and in nursing homes/nursing departments these parameters. [4,20]. It also justifies the continuous attempts to reevaluate, The 1 year mortality rate was 45% in the control group and modify and refine Beers' criteria and expand them to include 21% in the study group (P < 0.001). The annual referral rate to community-dwelling elders as well [3,17-19]. acute care facilities was also significantly lower in the study The updated Beers criteria may serve as an alarm system to group as compared to the control group (11.8% vs. 30% respec- increase physician alertness and avoid specific drugs in nursing tively, P < 0.002). homes/departments. We suggest that not only should we be There was an overall decrease in the cost of drugs in all aware of the high incidence of specific drug-related problems, departments. This improvement was represented by a $0.26 but we should thoroughly reevaluate the indications for each decrease in the average daily cost of drugs per patient in 132 drug. In this subpopulation, the sum total of the negative ef- patients in the four control departments (from $1.65 before to fects of a variety of drug combinations may outweigh the sum $1.39 after the intervention period). This change did not reach total of beneficial effects of the specific drugs. While comparing statistical significance (P = 0.07). However, a statistically signifi- risks versus benefits of drug withdrawal in this subpopulation, cant decrease of $0.46 in the average daily drug cost per patient one should remember that the rate of drug interactions is age- (from $1.74 to $1.28, P = 0.02, chi-square test) was shown in related, the odds of inappropriate medication use are higher as 190 patients in the six study nursing departments following the the absolute number of medications prescribed increases, and intervention (119 patients in the study group + 71 patients in the risk of hospitalization secondary to inappropriate medica- whom no change in drugs was made).
tion use is much greater in these facilities than in the general population [1,2,11,21]. Furthermore, the validity of indications and benefits of specific medications in this subpopulation is not There is an alarming increase in the number of people who suffer from disabling, non-curable diseases, which create exponentially Most guidelines for treating human maladies represent good increasing medical, economic and social age-related problems evidence-based medicine in middle-age patients. However, they [5]. The more years a person lives, the more age-related diseases may be inappropriate, with greater risks and lower benefits, will be acquired and the more drugs consumed. Polypharmacy, for institutionalized patients [22]. A well-accepted indication an age-related "geriatric syndrome," is a significant predictor of in adults may be unclear, no longer in existence, or irrelevant malnutrition, hospitalization and nursing home placement; it in the elderly, particularly in nursing facilities. For example, a impairs mobility and leads to morbidity and death [6]. patient who has received an antihypertensive or nitrates when For professionals in palliative medicine and particularly those still independent and active may not need these drugs years working in hospices, stopping drugs other than those used for later when already disabled and exerting minimal physical effort. symptom control is obviously a common practice. Nevertheless, Patients may also have a life expectancy that is shorter than the polypharmacy represents a problem also in palliative care settings time needed to benefit from any specific drug prescribed. [7,8]. However, in geriatrics, there is less awareness and attempts A similar approach has been adopted for disabled elders with to combat polypharmacy are much less aggressive.
diabetes [23]. In the absence of proven data for determining D. Garfinkel et al.
• Vol 9 • June 2007
Original Articles optimal glycemic control in frail elders, a panel of experts made number of patients in whom hypoglycemic medications were recommendations based on clinical judgment only. For the frail discontinued, statistical analysis was not relevant. elderly, those with short life expectancy and others in whom Globally, physicians are increasingly exposed to patients the risks of intensive glycemic control outweigh the benefit, suffering from a complexity of non-curable diseases. Nursing the panel did not adopt the general recommendations of the home/department patients may be treated by specialists who American Diabetes Association for lowering HbA1C to 7% and may work there part time while devoting the bulk of their time suggested a less stringent target of 8%. This approach should be elsewhere, or by less costly non-specialists, who usually represent expanded to include other clinical guidelines in an attempt to be the preferred choice of the nursing home/department manage- less aggressive in reaching rigid target goals (for example, blood ment. These patients may be taking medications that might have pressure, serum lipid concentrations), focusing rather on quality been given at some point in their lives by physicians of different of life and patient/family preferences. In line with this perception, specializations who prescribed the medication for a specific our approach aims at improving the quality of care in all 190 problem in their field of expertise. However, when policies were patients in the nursing department by reducing polypharmacy. determined by specialists, the nursing department physician may We have proven our hypothesis that several widely used types be reluctant to discontinue drugs even when a long time has of drugs are not necessarily needed in nursing home or nursing elapsed, new problems or medications accumulated, or physical department patients [Table 3]. changes occurred in the patient. Sometimes, neither special- Primum non nocera, our second hypothesis, was that our inter- ists nor the nursing physicians review all drugs in a search for vention would not have deleterious outcomes. Our findings that interactions with drugs prescribed by other doctors; therefore, both mortality and referrals to hospitals decreased significantly in a scheduled, formal drug reevaluation like ours may never be the study group are intriguing. The explanation that these find- ings are bias-related seems unlikely. Based on clinical judgment We have chosen the term "geriatric-palliative" to describe our only, physicians in the nursing departments decided whether to methodology for combating polypharmacy, because it is based re-administer drugs or send patients to an acute care facility. For on premises in both fields. All our patients suffer from non-cur- reasons of good medical practice, some patients in the study able diseases [Table 1] and our main goal is to relieve suffering group were monitored more frequently than those in the control using good palliative care medicine. The risk of polypharmacy group (e.g., more blood pressure assessments, ECGs, laboratory may outweigh the combined benefits of all drugs, and drug tests). However, as this mainly occurred in the first weeks and discontinuation in itself should be regarded as one of our high- the study was 12 months long, it does not explain the significant est therapeutic priorities. At least in this subpopulation, the annual differences in favor of the study group. well-accepted geriatric guideline "start low, go slow," should be Avorn and co-workers [22] concluded that drug discontinuation changed to "stop most, reduce dose." should be done selectively, altering one drug at a time. However, In the USA, for every dollar spent on medications used in in nursing department patients, who have the shortest life expec- nursing homes, $1.33 is spent to manage drug-related problems tancy and the worst quality of life, time is critical and they may [24]. Apart from the medical benefits, the financial benefits of suffer further deterioration due to drug-related problems from the our geriatric-palliative approach are considerable. Although it remaining medications. We therefore chose to withdraw several was performed in only 63% of patients in the study departments, drugs simultaneously, while carefully monitoring for any clinical the saving was still more pronounced than in the four control or laboratory adverse effects.
departments. Using this minimal estimate after correcting for Our study was not a randomized control trial. Nevertheless, the general saving represented by the control departments, the it provides evidence for the efficacy of our geriatric-palliative annual savings resulting from our approach would be $69 per approach. We recommend that randomized control trials be de- patient. This estimate is much lower than that found by Trygstad signed to conclusively assess our approach. However, performing et al. [12], who showed a relative annual cost reduction of $228 such trials on multi-drug discontinuation in the complex nursing per patient. Suppose our approach or that of Trygstad et al. was department/home subpopulation, while adhering to traditional implemented in at least 1.5 million nursing home patients in the rules of such trials, may be neither practical nor ethical. For USA and assuming the same cost of drugs, we would be looking example, it would require not only discontinuation of three to at an annual saving of 103 to 343 million dollars in the U.S. four specific drugs with no change in these same drugs in a alone, not including hospitalization savings.
comparable control group, but also continuation of the same Although the average number of medications consumed by drugs that are not withdrawn in both study and control groups. our patients was comparable to that reported by others, one One may argue that we have not provided direct evidence for a may argue that the success of our approach stems from the fact higher rate of drug-specific problems in the control group (e.g., that our patients were inadequately treated before the study. The higher incidence of orthostatic hypotension or hypoglycemia situation may be better or worse in other countries or specific in patients taking an antihypertensive or oral hypoglycemic, nursing homes/departments [25], but we believe that the extent respectively). However, orthostatic hypotension is not relevant of the problem is a global one. Therefore, using our approach to in disabled patients who cannot stand up. As for a possible confront polypharmacy can help improve the health of patients beneficial decrease in hypoglycemic events, due to the small and economies all over the world. In any case, the methodol- • Vol 9 • June 2007
Resolving Polypharmacy in the Disabled Elderly Original Articles ogy adopted can provide a useful checklist for even the best 13. Hanlon J, Schmader K, Rubi C, et al. Suboptimal prescribing in older inpatients and outpatients. J Am Geriatr Soc 2001;49:200–9.
14. Morton AH. Inappropriately defining "inappropriate medication for the elderly." J Am Geriatr Soc 2004;52:1580.
15. Beers MH. Explicit criteria for determining potentially inappropri- 1. Prescription Drugs and the Elderly: many still receive potentially ate medication use by the elderly: an update. Arch Intern Med harmful drugs despite recent improvements. Publication GAO/ 1997;157:1531–6. HEHS-95-152. Washington, DC: United States General Accounting 16. Chutka DS, Takahashi PY, Hoel RW. Inappropriate medications for Office; 1995:1–30. elderly patients. Mayo Clin Proc 2004;79:122–39. 2. Lau DT, Kasper JD, Potter DE, et al. Hospitalization and death 17. Fick DM, Cooper JW, Wade WE, et al. Updating the Beers criteria associated with potentially inappropriate medication prescriptions for potentially inappropriate medication use in older adults. Arch among elderly nursing home residents. Arch Intern Med 2005;165: Intern Med 2003;163:2716–24.
18. Goulding MR. Inappropriate medication prescribing for elderly 3. Aparasu RR, Mort JR. Inappropriate prescribing for the elderly: ambulatory care patients. Arch Intern Med 2004;164:305–12.
Beers criteria-based review. Ann Pharmacother 2000;34:338–46.
19. Zhan C, Correa-de-Araujo R, Bierman AS, et al. Suboptimal pre- 4. Gurwitz JH, Field TS, Avorn J, et al. Incidence and preventability scribing in elderly outpatients: potentially harmful drug-drug and of adverse drug events in nursing homes. Am J Med 2000;109:87– drug-disease combinations. J Am Geriatr Soc 2005;53:262–7. 20. Dhall J, Larrat EP, Lapane KL. Use of potentially inappropriate 5. Garfinkel D. Geriatric Boom Catastrophe – a major medical, drugs in nursing homes. Pharmacotherapy 2002;22:88–96.
economic and social nightmare of the 21st century. Proceedings 21. Zhan C, Sangl J, Bierman AS, et al. Potentially inappropriate of the 16th Congress of the International Association of Geron- medication use in the community dwelling elderly: findings from tology, 1997:364.
the 1996 Medical Expenditure Panel Survey. JAMA 2001;286:2823– 6. Frazier SC. Health outcomes and polypharmacy in elderly indi- viduals: an integrated review. J Gerontol Nurs 2005;31:4–11.
22. Avorn J, Gurwitz JH. Drug use in the nursing home. Ann Intern 7. Hanks G, Roberts CJC, Davies AN. Principles of drug use in palliative medicine. In: Doyle D, Hanks G, Cherny N, Calman K, 23. California Healthcare Foundation/American Geriatric Society Panel eds. Oxford Textbook of Palliative Medicine, 3rd. edn. New York: on Improving Care for Elders with Diabetes. Guidelines for im- Oxford University Press, 2004:214–25.
proving the care of the older person with diabetes mellitus. J Am 8. Twycross RG, Bergle S, John S, Lewis K. Monitoring drug use in Geriatr Soc 2003;51:S265–80.
palliative care. Palliat Med 1994;8:137–43.
24. Bootman JL, Harrison DL, Cox E. The healthcare cost of drug- 9. Gurwitz JH, Soumerai SB, Avorn J. Improving medication prescrib- related morbidity and mortality in nursing facilities. Arch Intern ing and utilization in the nursing home. J Am Geriatr Soc 1990; 25. Fialova D, Topinkova E, Gambassi G, et al. Potentially inappropri- 10. Beers MH, Ouslander JG, Fingold SF, et al. Inappropriate medica- ate medication use among elderly home care patients in Europe. tion prescribing in skilled-nursing facilities. Ann Intern Med 1992; 11. Cooper JW. Adverse drug reaction-related hospitalizations of nurs- ing facility patients: a 4-year study. South Med J 1999;92:485–90. Correspondence: Dr. D. Garfinkel, Head, Dept. of Evaluation &
12. Trygstad TK, Christensen D, Garmise J, et al. Pharmacist response Rehabilitation and Palliative Unit, Shoham Geriatric Medical Cen- to alerts generated from Medicaid pharmacy claims in a long- ter, Pardes Hana 37000, Israel. term care setting: results from the North Carolina polypharmacy Phone: (972-4) 637-566; Telefax: (972-4) 637-5757 initiative. J Manag Care Pharm 2005;11:586–7.
Serological testing for celiac disease
Hopper and co-workers tried to determine an effective di- The prevalence in the high risk and low risk groups was agnostic method of detecting all cases of celiac disease in 9.6% and 0.5%. The prevalence of celiac disease in patients patients referred for gastroscopy without performing routine who were negative for tissue transglutaminase antibody was duodenal biopsy. An initial retrospective cohort of patients 0.4% (7/2000). The sensitivity, specificity, positive predictive attending for gastroscopy was analyzed to derive a clinical value and negative predictive value for a positive antibody decision tool that could increase the detection of celiac result to diagnose celiac disease was 90.9%, 90.9%, 28.6% disease without performing routine duodenal biopsy. The and 99.6%, respectively. Evaluation of the clinical decision tool incorporated serology (measuring antibodies to tissue tool gave a sensitivity, specificity, positive predictive value transglutaminase) and stratifying patients according to their and negative predictive value of 100%, 60.8%, 9.3% and 100%, referral symptoms. No cases of celiac disease were missed by the pre-endoscopy testing algorithm. The prevalence of Br Med J 2007;334:729 celiac disease in patients attending for endoscopy was 3.9%. D. Garfinkel et al.
• Vol 9 • June 2007

Source: http://drugstop.co.il/articles/drugs/GARFINKEL'SPROCEDUREinNursingDepartments.pdf