Abstract
Background:
The effects of polypharmacy and its related adverse outcomes are well documented among elderly patients. Deprescribing is now recognised as an important part of medication review in addressing polypharmacy. However, little is currently known about local practices in discontinuing medications as a means to improve quality of life. There are two aims in this study: first, to explore the attitudes and beliefs of deprescribing medications among doctors in the Department of Internal Medicine (DIM) in Singapore General Hospital (SGH), and second, to see if differences exist among junior and senior doctors in their attitudes towards deprescribing.
Methods:
A descriptive survey was designed to look at the attitudes and beliefs of doctors towards deprescribing. All doctors in the department were invited to complete the survey.
Results:
The majority of doctors (66%) believed that deprescribing is beneficial for patients. Commonly cited barriers towards deprescribing were unwillingness to stop medications prescribed by another doctor, lack of time and insufficient knowledge. Lack of a specific approach to deprescribing was noted among 73% of the participants. A third of the juniors (32%) stated that they were reluctant to deprescribe medications, which was in contrast to the responses by the seniors, most of whom (94%) indicated no reluctance in deprescribing medications.
Conclusion:
Most doctors believe that deprescribing is beneficial to patients. Specific measures to target the barriers faced by doctors in deprescribing medications are necessary to improve deprescribing rates and minimising polypharmacy.
Keywords
Introduction
In developed nations, approximately one-third of elderly patients are prescribed five or more medications (defined as polypharmacy) and the number of prescribed medications usually increases with age. 1 Studies looking at the prevalence of polypharmacy in Singapore showed that at least half of all discharged patients from a Singapore restructured hospital had more than five chronic medications and about 30% were discharged with more than eight chronic medications. 2 The number of medications a patient takes is the single most important predictor of harm. 3 Elderly patients are at risk of undesirable consequences of polypharmacy which include increased risk of hospitalisation, falls and fractures, and even death, due to altered pharmacokinetics and pharmacodynamics.4,5
Deprescribing is the systematic process of reviewing, identifying, altering and discontinuing medications when existing and potential adverse outcomes outweigh the benefits. This practice may lead to amelioration in health outcomes within the context of an individual patient’s care goals, current level of function, life expectancy, values and preferences. 4 Medication review, including deprescribing inappropriate or unnecessary medications should be part of routine prescribing practice in all healthcare settings, especially in older populations, given that problematic prescribing is estimated to occur in more than one-third of patients over the age of 70. 6
There is now extensive evidence in the literature that deprescribing is not only feasible and safe, but more importantly, beneficial to patients. A systematic review of 31 withdrawal trials of specific classes of drugs in elderly patients above 65 years of age showed that the use of antihypertensive drugs, psychotropic drugs and benzodiazepines could be discontinued without harm in most patients with appropriate patient selection, careful withdrawal and close monitoring. 7 In a controlled trial conducted by Garfinkel et al. involving 190 patients in nursing homes in Israel, medications were successfully discontinued in 82% of the patients. 8 Additionally, the 1-year mortality rate was lower in the study group (21%) compared with the control group (45%) and there were also less referrals to acute care facilities in the study group.
A survey of Vancouver family physicians showed that most doctors were keen to deprescribe medications. However prescription by another healthcare specialist was identified as a barrier towards deprescribing medications. 9 In our local healthcare setting, patient care is fragmented and patients visit multiple specialists. 10 Also, there is little communication between primary care doctors and specialists in hospitals. This leads to a prescription cascade, whereby the side effects of medications are misdiagnosed as symptoms of another problem, resulting in further prescriptions and further side effects. 11 Studies showed that the process of reviewing and stopping unnecessary or potentially harmful medications is not yet part of routine practice. 12 Patients are seen by doctors with various levels of experience and studies show that junior doctors may not possess sufficient competence and confidence in deprescribing medications. 13
While it is clear that healthcare professionals play an important role in deprescribing, it is important to keep in mind that deprescribing is ultimately a patient-centred process. 4 The beliefs and attitudes of patients help to facilitate the process of deprescribing.4,12,14,15 A recent study by Ng et al. on patients attending public primary healthcare clinics in Singapore showed that the majority of patients were keen to reduce the number of medications they were taking and were willing to stop medications if their physicians thought it was no longer required. 15 While this study investigated the attitudes of patients towards deprescribing, 15 there is little information available in the literature on the attitudes of doctors in the local Singapore context.
The Department of Internal Medicine (DIM) is the largest clinical department in Singapore General Hospital (SGH) and manages around 300 patients a day with approximately 34 senior doctors and 100 junior doctors doing their rotation at a time. The majority of patients admitted to the department are elderly. Both senior and junior doctors play a pivotal role in medication review and deprescribing medications. Our study aims to understand the attitudes and beliefs of junior and senior doctors rotated to DIM towards deprescribing. We also further analysed the barriers doctors face when deprescribing and propose interventions to improve current practices to reduce polypharmacy in Singapore.
Methods
Study design and survey
A descriptive study was designed to understand the views, attitudes, and challenges faced in deprescribing medications in the department. The survey was designed by a focus group consisting of pharmacists and doctors based on previous studies done among doctors in Vancouver, Canada and the United Kingdom.9,13,16 The survey tool was then piloted among a small group of doctors to determine the time required to complete the survey and the ease of comprehending and responding to questions. General feedback on the survey design and questions were solicited. The feedback was discussed in various research meetings and the survey was edited accordingly. A copy of the final version of the survey is provided in Appendix 1.
Participants and administration of survey
Eligible participants were defined as doctors practicing in our department for at least 1 month. House officers, medical officers and junior residents at different years of training were categorised under junior doctors. Senior residents, registrars and consultants were categorised under senior doctors. Doctors eligible to complete the survey received an email with a link to the online survey. A total of 34 senior doctors and 200 junior doctors were invited to complete the survey. Survey responses were collected in January 2016 and January 2017.
Analysis
Descriptive statistics were used to report the collated survey and plotted as histograms. Due to the small sample size, no inferential statistics were applied.
Results
A total of 34 senior doctors and 56 junior doctors completed the survey voluntarily with a response rate of 100% among senior doctors and 28% among junior doctors.
When asked to rank medications that the doctors commonly deprescribed, the most common class of medication ranked first was antihypertensives (34.4%), followed by vitamin/supplements (17.8%) and opioids (12.2%) (Table 1). Medication classes that were ranked second included proton pump inhibitors (PPIs) at 21.1%, vitamin/supplements at 13.3%, as well as antihypertensives and statins both at 12.2%. Vitamin/supplements (13.3%), antihypertensives (12.2%), PPIs (12.2%) and benzodiazepines (12.2%) were the common medication classes that were ranked as third. Overall, antihypertensives, followed by PPIs and vitamins/supplements were the top three commonly deprescribed medications regardless of rank (Figure 1) and the responses were similar between junior and senior doctors.
Medications commonly deprescribed by DIM doctors by rank and their corresponding frequencies (%).
DIM: Department of Internal Medicine in Singapore General Hospital.

Medications commonly deprescribed by doctors rotating to Department of Internal Medicine, Singapore General Hospital.
The majority of doctors (66%) believed that deprescribing is beneficial to patients, with senior doctors more likely to make it a point to intervene for suitable patients as compared with junior doctors (91% versus 73%). The top three reasons for deprescribing were to reduce harm in view of side effects, followed by reducing pill burden for patients and minimal benefit of medications to patients (Figure 2). With regards to the frequency of deprescribing medications, 42% of the doctors indicated that they deprescribed medications a few times a week, 29% at least once a week and only 11% on a daily basis. However, the majority of doctors (73%) did not have a specific approach towards deprescribing.

Reasons for deprescribing medications by rank.
A third of the junior doctors (32%) stated that they were reluctant to deprescribe medications, which was in contrast to the responses by the senior doctors, where a vast majority of them (94%) indicated no reluctance. This may be attributed to the fact that the juniors think that they either do not possess (36%) or are unsure if they possess (39%) sufficient knowledge to deprescribe medications. On the other hand, 68% of the senior doctors believe that they have sufficient knowledge to perform this process.
The challenges to deprescribing medications are summarised in Figures 3(a) and 3(b). Both groups of doctors faced the biggest challenge in deprescribing when medications were prescribed by another doctor. Time constraints were cited as another major reason by both seniors and junior doctors. Junior doctors were more concerned about having inadequate knowledge in deprescribing, while senior doctors worried about resistance from patients and/or family members and adverse events from deprescribing.

Barriers faced by senior doctors in deprescribing medications by rank.

Barriers faced by junior doctors in deprescribing medications by rank.
The factors that enable deprescribing are shown in Figures 4(a) and 4(b). Junior doctors ranked having specific guidelines as the most important factor that would enable them to deprescribe medications, while senior doctors ranked having a pharmacist in the medical team as the most important factor. Recommendations by pharmacists were also among the top three factors that facilitate deprescribing of medications.

Factors that enable deprescribing of medications as ranked by senior doctors.

Factors that enable deprescribing of medications as ranked by junior doctors.
Discussion
The majority of doctors rotated to DIM who participated in this survey understood the benefits of deprescribing medications to patients. Not surprisingly, seniors were more likely to make it a point to deprescribe medications compared with junior doctors. Junior doctors are usually the first point of contact with patients upon admission and part of their job scope involves ordering and reviewing medications for these patients. However, in our study, junior doctors in general feel that they do not possess sufficient knowledge to deprescribe medications for patients and this is consistent with findings from other studies.13,14 This finding could be due to the junior doctors’ belief that medication review is the responsibility of the senior doctors. 13 One way to overcome this barrier is to implement an educational approach that equips juniors with the knowledge to deprescribe. 13 The inclusion of relevant material in the undergraduate and postgraduate curriculum will also help to change their perception and increase the confidence towards deprescribing from an early stage in their medical career.
In this survey, deprescribing medications prescribed by another doctor was the most cited challenge which is consistent with the findings of the Vancouver study. 9 The rapid growth in the number of specialist and subspecialist departments in Singapore has resulted in fragmentation of patient care with multiple doctors being involved in individual patient care. 10 Minimal communication between doctors, inadequate transfer of information at care interfaces and difficulty accessing medical records pose challenges to deprescribing.14,17 Doctors may not know the original indication for the medication and may find it too time consuming to consult the original prescriber.9,14,17 Although damaging relationships with the original prescriber was infrequently cited by participants in our study, a frequently mentioned barrier in other studies was a reluctance to question another colleagues’ prescribing decision associated with the respect for professional judgement and autonomy. 14 Due to the hierarchy that exists in the medical field, junior doctors may hesitate to question a more senior colleague to find out the original indication of a medication.13,14,17
Many participants also cited time constraints as a barrier for deprescribing and this is consistent with findings from other studies.9,14 A heavy daily inpatient load poses a critical challenge in performing a medication review. Having a pharmacist embedded in the medical team to suggest possible medications to be deprescribed may be useful in these instances. 16 This was cited as one of the top factors for enabling doctors to deprescribe medications in this survey, suggesting the need for a close collaborative relationship between doctors and pharmacists to optimise medication regimens and minimise adverse consequences of polypharmacy for patients. 18
Interestingly, the seniors cited resistance from patients and their family members as a barrier for deprescribing. A local study performed in a primary care setting concluded that high physician trust score as a significant factor in influencing patients’ attitude towards deprescribing. 15 Continuity of care by the same physician helps to enhance physician trust.15,19,20 Unfamiliarity with the medical team during hospitalisation may lead to resistance among patients and family members when deprescribing. Patients may be afraid of the adverse events after stopping a medication and are likely to be more receptive to deprescribing when they are assured that a discontinued medication can be restarted if necessary. 12 Patients who are concerned about the side effects of their medications and those who dislike taking multiple medications are also more willing to have their medications deprescribed by their doctors. 19 Studies also show that primary care doctors can be influential in encouraging patients to discontinue potentially harmful medications.12,14,15 After the medication is discontinued, a follow-up visit with either the primary care doctor or the inpatient doctor should be arranged to monitor for possible adverse events. Hence, deprescribing should be patient-centred and healthcare providers should engage patients and their caregivers regularly.
Study strengths and limitations
This is the first interdisciplinary collaborative local study surveying the perceptions of doctors towards deprescribing to the authors’ knowledge in SGH. There is currently little research available in the literature on doctors’ attitudes towards deprescribing in the local Singapore context. This study aims to understand the challenges faced by doctors towards deprescribing and therefore facilitates the development of strategies to address them. In addition, the responses between junior and senior doctors were differentiated in order to tailor strategies towards the needs of the doctors.
Despite its contribution, this study has limitations that should be addressed in further research. The survey participants were primarily involved in inpatient care and therefore the findings are mainly limited to doctors working in inpatient care in tertiary hospitals. Nevertheless, SGH is the largest tertiary hospital in Singapore and about 30% of the patients in SGH are admitted under DIM. Further research should include primary care doctors and doctors serving community hospitals and aged-care facilities.
Other limitations of the study included a small sample size, over-reliance on self-reported practices and a low response rate among junior doctors, possibly resulting in an increase in non-response bias. The true practices may actually vary from what was reported and more research is needed to explore the attitudes, beliefs and knowledge of junior doctors in order to develop further strategies for deprescribing.
Future directions
In response to the survey results, an algorithm on deprescribing of antihypertensives, PPIs and vitamin/supplements was devised by a group of DIM doctors and pharmacists. This algorithm serves as a guideline to both DIM junior doctors and pharmacists on the considerations that need to be made before deprescribing these medications. In order to facilitate closer collaboration between doctors and pharmacists in deprescribing, pharmacists are encouraged to perform medication reconciliation, first by interviewing patients upon admission to identify presence of adverse events, followed by compliance issues to these medications, and finally to recommend deprescribing to the DIM doctors as appropriate, according to the algorithm. A study is currently ongoing to determine the feasibility of the deprescribing algorithm for these specific medications (unpublished data).
For deprescribing to be successful, there needs to be a change in mindset among healthcare professionals and patients. Patients should be familiar with the indications and side effects of their medications and should be encouraged to seek clarification with a health care professional when in doubt.4,9,12,14,17 Healthcare professionals should make deprescribing part of their daily medication review and should engage patients in the process.4,13,14 Medication review tools can enable junior doctors to identify inappropriate medications.13,21,22 Junior doctors should be encouraged to clarify the indication for these medications not only with seniors in their department, but also with the original prescriber. 13
The National Electronic Health Records (NEHR) should be enhanced to improve communication among health care professionals involved in the various interfaces of patient care. 23 It can serve as a platform for health care professionals to document the indication of medications, reasons for discontinuing medications and in alerting other health care professionals about the use of potentially inappropriate medications. This will also address the challenge of deprescribing medications prescribed by other doctors.
Conclusions
As the elderly population in Singapore continues to increase, there needs to be more measures taken to reduce the harmful effects of polypharmacy. This study highlights the fact that most DIM doctors believe that deprescribing medications can be beneficial to patients, and a large proportion of them make it a point to deprescribe medications in their daily work. However, junior doctors still face various challenges in deprescribing medications. Apart from embedding pharmacists in medical teams, developing proper guidelines and promoting education to improve awareness and knowledge around deprescribing would enable doctors to deprescribe medications more effectively.
Footnotes
Appendix 1
House officer
Medical Officer
Senior Resident/Registrar
Associate Consultant
Consultant
Senior Consultant
Benzodiazepines
Antidepressants
Antiplatelet
Antianginal
Antipsychotics
Antihypertensive
Proton Pump Inhibitors
Statin
Bisphosphonates
Cholinesterase inhibitors
Anticonvulsants
Opioids
Vitamins/Supplements
Others : Please state__________________________
To reduce cost
To reduce harm to patient in view of side effects
To reduce pill burden
Because based on latest guidelines, the medication is not indicated
Because medication has minimal benefit for patient in view of age and comorbidities
Others: Please state__________________________
Lack of time to consider deprescribing
Do not have sufficient knowledge in deprescribing
Medications usually prescribed by another doctor and I’m unsure of the rationale
Concerned about adverse events after deprescribing medication
Damaging relationship with original doctor who prescribed medication
Resistance from patient/family
Unable to make patient understand rationale for deprescribing
Pressurized to prescribe according to guidelines
Others: Please state__________________________
Lectures on deprescribing specific medications
Guidelines on deprescribing specific medication
Having a pharmacist in your team
Flags by pharmacist to deprescribe medications in a patient-centred approach
Strong department focus on deprescribing medication
Others : Please state__________________________
Acknowledgements
We thank Xin Xiao Hui from the Academic Clinical Programme for Medicine, Singapore General Hospital for editing our survey and Foo Yang Yann from the Duke-NUS Medical School for her suggestions in revising our manuscript.
Conception or design of the work was provided by: Krithikaa Nadarajan, Jie Lin Soong, Tharmmambal Balakrishnan and Mei Ling Yee.
Data collection was provided by: Jie Lin Soong, Krithikaa Nadarajan, Tharmmambal Balakrishnan and Mei Ling Yee.
Data analysis and interpretation was provided by Jie Lin Soong.
Drafting the article was done by: Krithikaa Nadarajan and Jie Lin Soong.
Critical revision of the article was undertaken by: Jie Lin Soong, Tharmmambal Balakrishnan and Mei Ling Yee.
The final approval of the version to be published was done by: Krithikaa Nadarajan, Jie Lin Soong, Tharmmambal Balakrishnan and Mei Ling Yee.
The data sets used and analysed during the current study are available from the corresponding author on reasonable request.
Declaration of conflicting interest
The authors declare that there is no conflict of interest.
Informed consent
Participation was voluntary and participants were informed of the purpose of the survey.
Ethical approval
Not applicable as the study did not involve patients.
Funding
The research received no specific grant from any funding agency in public, commercial, or not-for-profit sectors.
