Avoiding Adverse Drug Withdrawal Events When Stopping Unnecessary Medications According to the STOPPFrail Criteria

Abstract

OBJECTIVE:

To provide clinicians with information about avoiding adverse drug withdrawal events (ADWEs) when discontinuing unnecessary medications as per the STOPPFrail criteria.

DATA SOURCES:

Searches of MEDLINE (1970-June 2020), the Cochrane Database of Systematic Reviews (through June 2020), Google Scholar (through June 2020).

STUDY SELECTION:

Reviews and original studies of ADWEs.

DATA EXTRACTION:

Tapering protocols for specific drugs/ classes from randomized controlled deprescribing trials.

DATA SYNTHESIS:

Six drug classes were identified as being high risk for physiological ADWEs.

CONCLUSION:

The occurrence of ADWEs is rare in comparison to adverse drug reactions in older adults. Few drugs/classes have been reported to have physiological ADWEs with abrupt discontinuation. For these we provide information about tapering protocols and symptom monitoring to avoid ADWEs.

Introduction

It is well known that polypharmacy is a common phenomenon in older patients.^1-3 The seminal conceptual work by Holmes et al. alerts clinicians to consider the transition of medication indications at end of life from curative to palliative in older patients.⁴ At this transition point, the emphasis changes to optimize medications that may improve/maintain function and quality of life and minimize/discontinue medications whose indication is preventative.⁴

The question then is, what constitutes a drug that can be deprescribed? One such approach is the application of explicit criteria such as those developed for patients with severe dementia and cancer.^5,6 While these may be helpful for specific conditions, they leave one unsure of their universal application. To fill this void, a set of explicit criteria named the Screening Tool of Older Persons Prescriptions in Frail adults with limited life expectancy (STOPPFrail) was published in 2017 by Lavan and collegues.⁷ They are intended to be relevant to those patients with end-stage irreversible pathology, poor 1-year survival prognosis, severe physical functional impairment or cognitive impairment of both and those patients where symptom control is the priority rather than prevention of disease.⁷ The STOPPFrail criteria note that for this set of end-of-life patients, the approach to medication uses or starting new medications should include the following considerations: (1) risk of the medication outweighing the benefit; (2) administration of the medication is challenging; (3) monitoring of the medication effect is challenging; (4) drug adherence/compliance is difficult; (5) lack of clear indication; and (6) intolerance of medication. We believe that an additional consideration should be whether the drug is high risk for physiological adverse drug withdrawal events (ADWEs). Since then, to the best of our knowledge, only 1 randomized controlled trial by Curtin et al. has targeted an intervention to reduce questionably beneficial medications per the STOPPFrail criteria.⁸ The appendices of the manuscript by Curtain et al. gave advice regarding tapering methods for medications identified by the STOPPFrail criteria. However, they are limited by the lack of specificity and supporting citations, particularly for those discontinued medications that have an increased risk of physiological ADWEs that represent the return of the underlying disease. Given this background, this review aims to provide clinicians with information about avoiding ADWEs when discontinuing unnecessary medications per the STOPPFrail criteria.

Methods

Articles that assessed ADWEs and deprescribing interventions in older adults were identified through searches of MEDLINE (1970-June 2020), the Cochrane Database of Systematic Reviews (through June 2020), and Google Scholar (through June 2020). Additional publications were identified by a manual search of the reference lists of identified articles, the authors’ own materials, published reviews, and Internet websites.

Data Synthesis

The literature search identified 6 seminal book chapters and reviews to identify those drugs when discontinued mentioned having physiological ADWEs.^9-14 We then reviewed primary ADWE study literature.^15-19 We then identified deprescribing randomized controlled trials to identify taper protocols.^20-32 Finally, we reviewed medication discontinuation algorithms that appear on national deprescribing organizations.^33-35

Table 1 lists those STOPPFrail drugs that when discontinued are unlikely to result in physiological ADWEs. This differs slightly from those listed in the appendix of the Curtin et al. of possible ADWEs, which includes angiotensin-converting enzyme inhibitor/ angiotensin receptor blockers, estrogen, nutritional supplements, oral diabetic agents, or theophylline, drugs for which we could find no evidence from a literature review that when these drugs discontinued cause physiological ADWEs.⁸

Table 1.

STOPP-Frail Criteria Drugs without Physiological Withdrawal Reactions

Drugs/Classes
Angiotensin converting enzyme inhibitor to prevent diabetes nephropathy
Antibiotics, prophylactic
Alpha blockers in catheterized patients (i.e., terazosin, prazosin, doxazosin)
Angiotensin receptor blockers to prevent diabetic nephropathy
Antiplatelets
Bisphosphonates and other antiresorptives
Calcium
Diabetes oral drugs
Estrogen for menopausal symptoms
Five (5) alpha-reductase inhibitors (i.e., dutasteride, finasteride)
Lipid lowering agents
Leukotriene antagonists
Memantine
Multivitamins
Nonsteroidal anti-inflammatory drugs taken long term
Nutritional supplements
Selective estrogen receptor modulators
Theophylline

Nonetheless, there are tapering algorithms for many of these drugs, including oral diabetes drugs as well as memantine on the Canadian deprescribing organization website, and for nonsteroidal anti-inflammatory drugs (NSAIDs), bisphosphonates, statins, and calcium from the Tasmanian deprescribing organization website.^33,34 Further, though generally thought that rapid discontinuation of peripheral alpha-blockers can cause rebound hypertension, data from benign prostatic hypertrophy clinical trials show that abruptly stopping peripheral agents (ie, terazosin, prazosin, and doxazocin) is feasible and safe.¹²

For the remaining drugs in Table 1, specific tapering guidelines are lacking. To prevent ADWEs resulting from the underlying disease, clinicians should consider the dose, duration, and pharmacokinetics of medications.⁹ Risk can be minimized or eliminated by slow, careful tapering of the medication over a prolonged period of time (eg, 4-6 weeks). This approach is similar to that taken in the initiation and titration of a new medication. Careful monitoring for early symptoms of the underlying disease (eg, knee pain after discontinuing NSAIDs) is advised.

Table 2 lists the medications that are part of the STOPPFrail criteria with an increased likelihood of physiological ADWEs. This table also list signs/symptoms seen with a typical physiological ADWE and tapering recommendations to avoid these physiological withdrawal reactions.

Table 2.

STOPPFrail Criteria Drugs with Physiological Withdrawal Reactions

Drugs/Classes	Typical Physiological Withdrawal Reactions	Tapering Recommendations to avoid Physiological Withdrawal Reactions
Central alpha blockers for hypertension (e.g., clonidine)	Rebound hypertension	Taper over 3 to 6 weeks28-29
Anticholinergic for colic symptoms or in those with urinary catheter	Cholinergic rebound symptoms including irritability, anxiety, insomnia, sweating and nausea	Taper over 4-8 weeks35
Antipsychotics used for behavioral and psychological symptoms of dementia	Dyskinesias, overactivity, super sensitivity psychosis	1st generation antipsychotic: 25%–50% daily dose reduction cholinergic every 1-2 days; 2nd generation antipsychotic: 25%–50% daily dose reduction every 2 weeks (except aripiprazole 25%–50% daily dose reduction every 2 months)36
Corticosteroids (oral) taken long term	Adrenal insufficiency including symptoms of nausea, fever, anorexia, lethargy, arthralgias, postural hypotension	Taper prednisone equivalent at a rate of 5 mg per week until reach 10 mg/day in prednisone equivalent. Then taper 0.5 mg-1 mg/day of prednisone equivalent every 2-4 weeks. May take 6-12 months to completely withdraw and see return of hypothalamic pituitary adrenal axis30,31
Histamine 2 receptor antagonists	Rebound gastric hypersecretion with symptoms of heartburn, reflux, dyspepsia, epigastric pain, weight loss	Half dose initially. Stop altogether in 1 month if no symptoms of dyspepsia33
Proton pump inhibitors	Rebound gastric hypersecretion with symptoms of heartburn, reflux, dyspepsia, epigastric pain, weight loss	Half dose initially. Stop altogether in 1 month if no symptoms of dyspepsia33

There is some literature that the abrupt withdrawal of the central alpha-blocker clonidine can cause rebound hypertension.²⁰ The average age of the 9 subjects studied was 62 years.²⁰ The mechanism is thought to result from increased peripheral sympathetic activity.²⁰ Data from antihypertensive deprescribing trials suggest a 3-to-6-week taper to be prudent especially when the patients is maintained on at least 1 other antihypertensive.^28,29

We found only three deprescribing trials of anticholinergics that measured associated health outcomes in older adults.^21-23 None showed consistent improvement in neuropsychological testing, nor did they specify a tapering schedule or report withdrawal reactions.^21-23 However, since cholinergic rebound is theoretically possible, the New South Wales deprescribing website offers tapering guidelines (ie, wean gradually by 25%-50% of the daily dose every 1-4 weeks).³⁵

Antipsychotics block a number of neurotransmitters, including dopamine, and thus rebound effects are possible.²⁷ None of the published deprescribing trials described an approach based on pharmacokinetic principles.^24,25 We recommend a 2-stage gradual dose-reduction protocol for tapering antipsychotic medications based on the log dose-response relationship; each stage was designed to result in 50% dose reductions prior to discontinuation, according to the specific half-life of each drug.³⁶ So, for example, someone taking long-term risperidone 4 mg daily whose half-life is 3-20 hours of the parent compound and 21-30 hours of the active metabolite, one would first decrease the dose to 2 mg daily and then 2-3 weeks later decreased it further to 1 mg daily and again 2-3 weeks later to 0.5 mg daily before discontinuation.

It is well known that abrupt discontinuation of systemic corticosteroids taken in high doses for long durations can result in signs of adrenal insufficiency.³⁰ One study included older subjects (average age older than 68 years) taking greater than 5 mg/day continuously for pulmonary problems. They recommended tapering prednisone equivalent at a rate of 5 mg per week until 6-11 mg per day was reached. A well-known drug information source further suggested tapering of 0.5-1 mg/day in prednisone equivalent occur every 2-4 weeks and monitor over 6-12 months.³¹

Abrupt discontinuation of H₂ blockers/proton pump inhibitors (PPI) has been reported to cause rebound acid hypersecretion and related symptoms.¹³ One study of 98 chronic PPI users without a history of peptic ulcer or esophagitis (median age 63 years) found that only 27% did not use PPIs during the year after discontinuation.³² The Canadian deprescribing website offers practical guidelines for tapering that includes cutting the daily dosage in half for 1 month and use as-needed antacids and then discontinue.³³

Discussion

To the best of our knowledge, this is the first review of ADWEs that might occur when discontinuing unnecessary drugs per the STOPPFrail criteria. It is important to note that this is not a comprehensive review of all possible ADWEs that can occur during deprescribing. In particular, serious physiological ADWEs can occur with abrupt discontinuation of antiepileptic drugs, baclofen, beta-blockers, and benzodiazepines, which are not part of the STOPPFrail criteria. Readers are encouraged to further review the following references for more information about withdrawing these drugs and others.^9-19

It is interesting to note that, with the exception of corticosteroids, that most drugs can be safely tapered over a 4-6 week period. While this time period might be too prolonged/complicated for some hospice patients, there is time for many hospice patients whose average length of stay is approximately 90 days.³⁹ Also, it is important to note that corticosteroids can be helpful for symptom control at end of life (eg, nausea and vomiting after palliative radiation) and should not always be considered unnecessary.

There are limitations to this review. It is not a systematic review and meta-analysis. In addition, the articles cited rarely included frail older adults. More studies such as the recent article by Niznik et al. are needed to determine the consequences (if any) of discontinuing medications at end of life.⁴⁰

Conclusion

The occurrence of ADWEs is rare in comparison with adverse drug reactions in older adults.^9-13,37,38 Few drugs/classes have been reported to have physiological withdrawal reactions with abrupt discontinuation.^9-13,37,38 For these we provide insights as to what symptoms to look for and information about tapering unnecessary medications to avoid ADWEs. Further empiric studies are needed to measure actual ADWE event rates of these protocols.

Addendum

Since the writing of this manuscript, version 2 of the STOPPfrail criteria has been published.⁴¹ There are new criteria for deprescribing antihypertensives, antianginals (ie, nitrates, nicorandil, ranolazine), and vitamin D.⁴¹ Regarding the first new criteria, recently reported were the results of a randomized controlled deprescribing trial of older adults taking multiple antihypertensives with a systolic blood pressure lower than 150 mm Hg.⁴² They found that one antihypertensive could be discontinued without significant adverse outcomes or increases in blood pressure. Caution should still be taken to taper alpha (eg, clonidine) and beta-blockers (eg, propranolol) over 3 to 6 weeks as physiological adverse drug withdrawal events have been reported.^43,44 Discontinuation of antiantiginals should not be attempted in those with chest pain in the previous 6-12 months or in those with known coronary artery disease.⁴¹ In those who have these antianginals deprescribed, 30 days of careful monitoring should be undertaken to look for relapse of anginal symptoms.⁴⁵ Finally, no particular precautions are necessary to abruptly discontinue vitamin D.

ABBREVIATIONS:

ADWE

Adverse drug withdrawal events

NSAIDs

Nonsteroidal anti-inflammatory drugs

PPIs

Proton pump inhibitors

STOPPFrail

Screening Tool of Older Persons Prescriptions in Frail adults with limited life expectancy