Drug-related problems in hospitalized patients with type 2 diabetes mellitus: A systematic review

Introduction Type 2 diabetes mellitus (T2DM) is one of the non-communicable diseases which continues to rise in prevalence and mortality rate throughout the years. Drug-related problems (DRPs) are more prevalent among T2DM patients especially those with co-morbidities. Objective The objective of this study was to review and assess the prevalence and characteristics of DRPs among hospitalized type 2 diabetes mellitus patients. Methods The systematic review of the literature was carried out using five online databases: PubMed, Scopus, Google Scholar, Web of Science, and Cochrane Library from the inception of the database until June 2022. Studies included in the review were published in English or Malay language. The data were extracted and assessed using the Joanna Briggs Institute (JBI) critical appraisal tools. Results A total of 939 studies were identified with 20 studies that met inclusion criteria and were included in this systematic review. The overall prevalence of DRPs in all 20 studies ranged from 7% to 94%. The most common DRPs included drug-drug interaction (DDI), adverse drug reaction (ADR), therapeutic effectiveness problems, and inappropriate medication use. Conclusion The most common drug classes involved were antidiabetics (metformin), antihypertensives, antiplatelets and antibiotics. The risk factors contributing to DRPs included the presence of comorbidities, the number of medications, and polypharmacy. To conclude, the rate of DRPs incidence in hospitalized T2DM patients was observed to be high. Further future studies with appropriate study designs and methods of detecting DRPs will be necessary to reduce and prevent DRPs occurrences.


Introduction
The administration of medications is essential for the treatment and prevention of health issues.Throughout the years, the number of medications developed is increasing which has led to the discovery of more complex drug regimens to treat patients.Despite the remarkable advance in drug development to treat patients, Drug-Related Problems (DRPs) may arise which can cause undesirable treatment outcomes and potentially worsen patient conditions.According to the Pharmaceutical Care Network Europe (PCNE), DRP is defined as "an event or circumstance involving drug therapy that actually or potentially interferes with desired health outcomes". 1There are nine sub-categories identified causes of DRPs: 1. Improper drug selection 2. Improper dose selection 3. Inappropriate dose selection 4. Duration of treatment 5. Problem with prescribing and dispensing process 6. Issues with drug administration process 7. Patient-related including behavior 8. Patient transfer related, between primary, secondary, and tertiary care. 1 The three main types of DRPs include Medication Error (ME) defined as "a failure in the treatment process that leads to, or has the potential to lead to, harm to the patient". 2,3Adverse Drug Reactions (ADR) defined as "any response to a drug which is noxious and unintended, and which occurs at doses normally used in humans for prophylaxis, diagnosis or therapy of disease, or for the modification of physiological function", 4 and Adverse Event defined as "an adverse outcome that can be attributed, with some degree of probability, to an action of a drug". 5he prevalence of T2DM increases with time and is globally estimated to be most common in people aged 55 to 59 years with males experiencing symptom manifestation sooner than females.In 2019, the number of adults diagnosed with T2DM was found to be the highest in China and India estimated with 116.4 million and 77.0 million, respectively.According to the IDF, the number of people with T2DM is expected to increase to 700 million globally. 6The first-line drug therapy for T2DM is generally metformin with several advantages including a lower risk of hypoglycemia and promoting weight loss.Depending on the patient individual glycemic needs, additional medications from different drug classes (eg., glucagon-like peptide 1 [GLP-1] receptor agonists, sodium-glucose cotransporter 2 [SGLT2] inhibitors), may be appropriate.8][9] Studies also confirmed that insulin therapy was associated with a higher incidence of ADR due to the hypoglycemic tendency of its use. 10revious systematic reviews focused on patients with diabetes' admission to hospital due to medication problems 11,12 with several reviews focused on patients with type 2 diabetes in an outpatient setting. 13,14While there are various systematic reviews focusing on the identification of DRP in patients with diabetes in a hospital setting, few systematic reviews have been conducted on the prevalence and risk factors in hospitalized patients specifically with type 2 diabetes.Therefore, the aim of this study was to review and assess literature related to DRP in adult hospitalized T2DM patients, focusing on its prevalence and risk factors.

Methods
The Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols was used for this systematic review.Five electronic databases were used for the review including PubMed, Scopus, Google Scholar, Web of Science, and Cochrane Library.The articles were searched from the inception of the database until June 2022.Various search key terms were used for each database including drug-related problem (DRP), adverse drug reaction (ADR), adverse drug event (ADE), medication error (ME), treatment related problems, medicalproblem oriented plan, inpatients, and type 2 diabetes mellitus (T2DM).
Studies on DRPs or subcategories in adult (≥ 18 years) type 2 diabetic hospitalized patients of both genders were included in the review.Type 2 diabetes without or with comorbidities (eg., Hypertension, renal failure) was included in the review.Type 1 diabetes and gestational diabetes were excluded from the review.However, the search was not limited to studies that report only type 2 diabetes.If a study consists of data on DRPs for both type 1 and type 2 diabetes patients, the study was read and data on type 2 diabetes were extracted.If data on type 2 diabetes were unable to be extracted, the study was removed.A similar selection process was applied to studies that consist of data on DRPs involving type 2 diabetes in both inpatient and outpatient settings.
The type of study designs reporting data and prevalence of DRPs in type 2 diabetes hospitalized patients included cross-sectional studies, interventional studies, and cohort studies.Studies included in this review were published in English and Malay.
A software program Endnote (Chandler, AZ, United States) was used for the systematic review where the study retrieved from the database search was transferred.The software facilitates the identification and removal of duplicate articles.Two researchers first screened the articles based on the titles and abstracts.Then, the full articles were screened according to the inclusion and exclusion criteria with reasoning provided for studies that were excluded at this stage.The discrepancies between two researchers were then independently reviewed by the third researcher.
Data pertaining to study details were extracted by two reviewers.Study details include the characteristic of the study (study design, duration, setting, patient demographics), DRPs characteristics (methods to classify DRPs, most common drugs involved, overall incidence), and the risk factors for DRPs.
The methodological quality of the studies included in the systematic review was assessed by two reviewers using the Joanna Briggs Institute (JBI) critical appraisal tools.The tool uses four answers: 'Yes', 'No', 'Unclear' and 'Not applicable'.
A total of 939 studies were identified from the five databases which subsequently 725 studies were excluded based on the title and abstract.A total of 88 studies were screened in full text whereby 68 studies were excluded for not fulfilling the inclusion criteria.Ultimately, 20 studies met inclusion criteria and were included in this systematic review.Fig. 1 shows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow chart.Characteristics of included studies and drug related problem are presented in Table 1 and Table 2, respectively.PRISMA checklist can be found in Supplementary Information 1.
A wide range of scores was allocated across the 20 studies with an overall of standard quality.Three studies used an inappropriate method (purposive sampling) of recruiting participants. 21,25,26Individual quality assessment scoring is presented in Supplementary Information 2.

Characteristics of studies included
Retrospective studies constituted the majority of the articles (40%), with different various observational methodological approaches (Table 1).Two prospective studies included were conducted with interventional methods implemented.Most of the studies were carried out in Asia (75%): Indonesia, 17,[21][22][23][24][25][26]28 India, 20,27,[30][31][32] Japan, 18 and Pakistan, 34 followed by Europe (15%): Italy, 15 Poland, 19 Austria, 33 Africa (5%) 29 and North America (5%). 16The majority of the studies were conducted in the general hospital ward with only 2 studies reported to have been conducted in the endocrinology and diabetology department. 19,33 In all o the 20 studies, a total of 6844 patients were recruited with only adults (≥ 18 years). Ofthese, five studies were focused only on geriatric patients (≥ 60 years).16,18,23,24 Two studies used surveys or questionnaires as means of collecting data on DRPs from the patient. 20,33Two studies do not provide information on methods of detecting DRPs.19,22 In all of the 20 studies, a total of 27 sources of information were used to detect and classify DRPs.Eight studies used one source only.18,20,21,25,29,31,32,34 Seven studies used more than one source.17,[22][23][24]27,28,30 One study establishes operational definitions as its source.6,33 Micromedex was the most used source of information (n = 3) 24,27,30 followed by Lexicomp (n = 2) 27,30 and American Diabetes Association (n = 2).22,28 The overall prevalence of DRPs in all 20 studies ranged from 7% to 94%.One retrospective study reported antidiabetic drug interactions only and found 6.12% of 49 patients experiencing DDIs.Three different sources were used to identify and classify the DRPs detected.
The majority of the studies investigated DRPs as a whole (n = 11) with no specification on the type of DRPs investigated, for example DDI, ADR, therapeutic effectiveness and dosage problems.Five of these studies identified DDIs as the main type of DRP found with a range of 29% to 94%. 17,20,27,28,31One retrospective study found that 37.6% of 250 patients have insufficient clinical information during the hospitalization period. 30The necessary clinical parameters for monitoring diabetes including blood glucose levels were not documented.Only one study reported therapeutic effectiveness problems as the common DRP with a frequency of 50.6%. 24The problems observed include the medication being ineffective and the failure of treatment.One study reported drug choice problems in 55.17% of 90 patients with the primary source of DRPs being drug/dose selection, which was similarly found in another study conducted by Nyazisenga and colleagues (2019).with frequencies of 62.16% and 36.97%respectively. 22,29No drug prescribed with clear indication has the highest frequency of 25.3% under the drug choice problem subdomain.
Two studies reported ADR only with one study focused on a specific antidiabetic drug (metformin) and another on a specific drug class (antihyperglycemic) with no drug name mentioned. 16,18In a study of ADR on a specific drug (metformin), the occurrence of diarrhea was found in 26.7% of 101 participants while a study of ADR on a specific drug class (antihyperglycemic) found 46.9% of 484 hypoglycemic episodes manifested in T2DM patients.
In this review, nine studies reported the most common drugs involved in DRPs, the majority of the drugs involved were antidiabetics, antihypertensives, antiplatelets and antibiotics.Four studies reported antidiabetics (metformin) to be commonly involved in DDIs. 20,23,30,31namdar and colleagues (2020) reported a common DDIs of metformin with metoprolol (33.1%) and ranitidine (26.4%). 31One study reported common DDIs with cardiovascular medications such as aspirin, clopidogrel, warfarin, and atorvastatin whereas antihypertensives such as hydrochlorothiazide and furosemide are associated with a higher occurrence ADRs. 273][34] One of the studies investigated a specific drug (metformin) use in the elderly and observed 82% of 335 patients to be contraindicated with only three reported dyspepsias during hospital stay. 19Two studies investigated PIM in the elderly population with a prevalence of 30% and 74% having at least one PIM. 21,32Two studies reported a ME prevalence of 25% having at least one ME and 70.3% respectively with the most common problem associated with dosage problems. 33,34The study with a 70.3% prevalence further specifies 29.2% receiving doses too high and 27.9% doses too low. 34These findings aligns with another study that reported up to average 9.35 treatment related problems per patient, with 53% considered as major and 28% as moderate treatment related problems in hospitalized internal medicine patients in Jordan. 35hree studies investigated DRPs in specific drugs or drug classes (metformin, antihyperglycemics). 16,18,19 Thirteen studies included the risk factors contributing to DRPs.The most prevalent risk factors for DRPs involved the presence of comorbidities found in six studies with the majority having hypertension (HTN). 20,22,27,28,31,34Five studies reported on age, 20,24,28,29,34 four studies on the number of medications, 22,[27][28][29] another four studies on polypharmacy, 20,29,31,32 and three studies on gender. 18,28,32The list of risk factors reported in the studies is presented in Supplementary Information 3.

Discussion
In this systematic review, we examine the overall prevalence of DRPs in T2DM hospitalized patients with risk factors influencing DRPs and the most common drugs involved with the DRPs.In this review, ten studies investigated DRPs as a whole, followed by six studies on ME, three studies on ADRs and only one study specifically on DDIs.
The ideal outcome of treatment for T2DM patients can be defined generally as attaining the targeted glycemic control mainly by the adjustments in lifestyle together with the initiation of drug therapy.Therapeutic success can be more thoroughly defined as slowing or halting disease development and minimizing the risk factors associated with disease complications. 36The first line of drug therapy for diabetes is the initiation of metformin with a slow intensification of therapy depending on the disease progression with the goal of achieving targeted glycemic control as stated previously. 7The presence of DRPs has a potential negative impact on the outcome of diabetes treatment as well as any other conditions.The consequences of DRPs interfere with the desired clinical outcome worsening patient quality of life which emphasizes the importance of pharmaceutical care in preventing unfavorable health outcomes. 37ne of the ways to diagnose diabetes is by using the A1C test to measure the average blood glucose for the past two to three months.A1C result showing a value of 6.5% and above indicates an individual is diagnosed with diabetes.Another test known as fasting plasma glucose (FPG) is performed during fasting when an individual is not consuming food or drinks for at least eight hours to measure the fasting blood glucose level.American Diabetes Association defines diabetes mellitus when an individual's fasting blood glucose value is 126 mg/dL and above. 38In this review, no studies specified the test used to diagnose patients with type 2 diabetes.Before the progression to T2DM, an individual often goes through a stage known as prediabetes in which they show a high level of blood glucose but are still not considered high enough to be identified as T2DM. 39he methods for identification of DRPs used in the studies include chart reviews, incident reports, surveys, direct observation, and minimal interviews.[25][26][27][28][29][30][31][32]34 The use of chart review at the prescription stage demonstrates a higher detection of DRPs with low clinical significance.The outcome of the chart review is dependent on the ability of the reviewer to detect the   triggers. 40,41The second most used method was incident reporting. 15,16,18,23,24Five studies implemented more than one method of detecting DRPs with combinations of chart review with direct observation, interview, or incident reports. 16,18,24,27,31,34n this systematic review, a wide variety of medications was involved with DRPs in all 20 studies.Most of the studies did not specify investigating one specific drug class only however, results presented in the studies were mostly associated with class of antidiabetic medications only.
Two studies performed an investigation on a specific drug which was metformin. 18,19Overall, the most common antidiabetic drug classes associated with DRPs in this review were biguanide (metformin), followed by cardiovascular drugs mainly antihypertensives with several studies specifically investigating DRPs in T2DM patients with HTN as comorbidity.The higher prevalence of comorbidities has greater exposure to DRPs as the presence of such complex conditions is often associated with multiple drug therapy, hence, the patient comorbid condition and therapy should be carefully reviewed ensuring appropriate prescribing of medications and early prevention of DRPs. 42ntihypertensives were found to be the most commonly involved in DRPs in three studies with one identifying the ACE inhibitor drug class mainly involved with DRPs.Multiple drug therapy regimens including antihypertensives with other cardiovascular medications potentially reduces the effect of therapy and have greater exposure to adverse effects. 43etformin demonstrates a significant involvement with DRPs in five studies. 20,23,30,31,34The DRPs involving metformin in these studies were mostly associated with DDIs with other non-antidiabetic drug classes including metoprolol, ranitidine, simvastatin, and ciprofloxacin.The mechanism of interaction between metformin with beta-blockers and H2 receptor blockers was mainly through inhibition of either organic cation transporters (OCT) or multidrug and toxin extruders (MATE) transporters or both.The reduction in metformin renal clearance and increase in metformin plasma concentration further increases the risk of developing metformin-associated lactic acidosis. 44,45Therefore, prescribers should be alert for patients with medications that have an effect on the OCTs and MATEs transporters.
Out of all twenty studies included, thirteen mentioned the risk factors contributing to DRPs.In all of the thirteen studies, a total of 23 risk factors were observed with the most prevalent being associated with the presence of comorbidities which increases number of medications reported in six studies with the majority having hypertension. 20,22,27,28,31,34The risk of developing hypertension in people with diabetes is higher as they often manifest insulin resistance compared to normal individuals. 46This was followed by the number of medications reported in four studies 22,[27][28][29] and polypharmacy in another four studies. 20,29,31,32Overall, the risk factor was mostly involved with the high number of prescribed medications resulting in higher exposure to DRPs. 47The exposure to DRPs can be reduced with the act of deprescribing ("a systematic approach to identify and discontinue medications in which potential harm outweighs the benefit and medications with unclear benefit"). 48

Strength and limitations
In the previous systematic reviews, the majority are focused on DRPs in the general inpatient diabetic population with a few on T2DM in inpatient and outpatient settings.This systematic review is one of the few that conducted reviews on the available literature on DRPs in hospitalized patients specifically with T2DM.Five electronic databases were used to search appropriate articles with various search strategies used in each database.The data on the most common drugs or drug classes involved with DRPs were limited.Several studies reported the name of common drugs involved only when it came to DDIs.The common drug or drug classes involved were stated without describing the DRPs it was associated with.Additionally, a few studies performed nonrandom sampling which may possibly generate highly biased data results.One of the studies that carried out a questionnaire study design based on patient perspective will have a risk of the individual not having the appropriate knowledge or recollection of the DRPs experienced to provide responses accurately affecting the results of the study.
From this systematic review, it can be observed that hospitalized T2DM patients have a high prevalence of DRPs with the most common being DDIs.Prescribers should be more cautious with this population, especially those who have co-morbidities as it increases the number of medications and risk of DDIs.The risk of DDIs could be reduced by regularly reviewing the patient list of medications and using a reliable source of information guidelines on common drug interaction with antidiabetics.Future studies on DRPs prevalence in hospitalized T2DM patients are necessary for further development of approaches in identifying, resolving, and preventing DPRs improving patient clinical outcomes. 49Recently, a new tool, The Alsayed_v1 tools, encompassing treatment assessments, medical-problem oriented plan, and care plan, have been validated and applied to real patient cases. 50This comprehensive tool can be modified for clinical pharmacists to efficiently identify, categorize, and address drug-related problems for T2DM.

Conclusion
In this systematic review, the occurrence of DRPs in T2DM hospitalized patients was observed to be high with a diverse range among the included studies.DDIs were reported to be most prevalent in a majority of the studies.Antidiabetics (metformin) and antihypertensives showed a significant involvement with DRPs.The risk factors that were highly associated with DRPs were the presence of comorbidities (HTN), the number of medications and polypharmacy.

Table 1
Characteristics of included studies.
(continued on next page) M.H. Awang Jihadi et al.

Table 2
Drug-related problem characteristics.