Assessment of Appropriate Prescribing in Maintenance Haemodialysis and Comparison of Prescribing Patterns Between State- and Privately-Funded Patients: A Retrospective Review

Abstract

Background:

Patients with end stage renal disease undergoing haemodialysis commonly have multimorbidity, which in turn leads to polypharmacy. Additionally, end stage renal disease limits the choice and dosage of medicines. In Namibia, patients can access care at privately run dialysis centres. Patients seen at state-run health facilities in need of renal care from the dialysis centres, have their renal care costs covered by the Ministry of Health and Social Services.

Aim:

The main purpose of this study was to assess the appropriateness of medication therapy among patients with end stage renal disease who underwent haemodialysis at two dialysis centres in Windhoek. Further, the study also compared prescribing patterns between state-funded and privately funded haemodialysis patients.

Methods:

A retrospective, quantitative and analytic design involving the review of clinical records of patients attending haemodialysis at the two dialysis centres. The types of medicines prescribed were classified according to the Anatomical Therapeutic Chemical (ATC) classification. The Drug Prescribing in Renal Failure Handbook was used as a primary reference for assessing the appropriateness of renal dosage adjustment. The primary references used to check for the appropriateness of medication choice were the KDIGO Clinical Practice Guidelines. Other references were used as secondary sources.

Results:

A total of 147 patients’ clinical records were reviewed and included in this study. About one-third (33.3%) of patients had at least one or more inappropriately selected medicines. Most patients 82.0% (n = 121) had at least one or more inappropriately renally adjusted medicines. More privately funded patients were prescribed Vitamin D or its analogues (P < .001), phosphate binders (P < .001), antithrombotic agents (P < .001), lipid modifying agents (P < .001), and angiotensin receptor blockers (P < .001).

Conclusion:

As anticipated, patients on haemodialysis were being managed with a relatively large number of medicines. Renal dosage adjustment could be improved to ensure patient safety. The differences in the prescribing of vital medicines such as phosphate binders, between state funded versus privately funded patients warrants further investigation.

Introduction

The number of people receiving dialysis globally is expected to increase to up to 10 million by the year 2030, with 70% of these individuals living in sub-Saharan Africa.^1,2 In many parts of the world, diabetes mellitus is the leading cause of end stage renal disease (ESRD); however, especially in low- and middle-income countries (LMICs), hypertension, systemic lupus erythematosus, and human immunodeficiency virus (HIV) infection contribute significantly. ³ The burden of chronic kidney disease (CKD) is also expected to increase significantly in Namibia due to the continued increasing prevalence of CKD risk factors such as hypertension, diabetes mellitus and HIV. For instance, in 2019, 52% of adults aged 30 to 79 years in Namibia were diagnosed with hypertension, yet only 22% of those on treatment had controlled blood pressure. ⁴

Patients receiving dialysis have a high burden of co-existing diseases, resulting in the use of many medicines and an increased risk of drug-related problems. Further, the pharmacokinetics of medicines may be significantly altered by ESRD.^5,6 Haemodialysis may effectively clear some small molecules from circulation, leading to subtherapeutic levels after a dialysis session. ⁷ These changes in drug pharmacokinetics therefore necessitate dosage adjustment, as failure to do so may lead to drug or metabolite accumulation, predisposing the patient to adverse effects. Appropriate medication selection and dosage adjustment are therefore crucial to prevent undesirable health outcomes, mortality, and morbidity in patients on renal replacement therapy.

Although several studies and systematic reviews have been conducted to assess prescribing in patients with CKD, only a few have specifically assessed those on haemodialysis.^8-17 These studies in haemodialysis patients were mainly conducted in high-income countries. Studies conducted in African countries^18-21 have mostly assessed prescribing in patients with various stages of CKD and not those on dialysis. One study conducted in Nigeria, ¹⁰ in which 53.7% of participants were on maintenance dialysis, reported only on drug-drug interactions and not on the overall appropriateness of therapy. This highlights a noticeable gap in the availability of data on the appropriateness of management of dialysis patients in sub-Saharan Africa, where unique challenges such as limited access to medicines and issues of medicine affordability persist.

The healthcare system in Namibia comprises of public and private sectors.^22,23 The public sector is financed by the national government under the Ministry of Health and Social Services (MOHSS) and caters for approximately 83% of the country’s population, predominantly people who are unable to pay for medical insurance or to pay out of pocket. ²³ At the time of this study, the MOHSS did not have any haemodialysis units, although some are now available. Therefore, in the event a state-funded patient required haemodialysis, the MOHSS paid for this service at a private dialysis centre. ²⁴ Although the World Bank classifies Namibia as an upper-middle-income country, Namibia has one of the highest rates of socioeconomic inequality in the world. ²⁵ According to the 2023 national census, 37% of the population was unemployed and 34% of employed people reported earning a gross income of USD$100 or less per month. ²⁶ As both private and state-funded patients receive dialysis from the same facilities, it is not known whether there are differences in prescribing patterns between patient groups.

Haemodialysis services, as well as medicines needed to manage complications of kidney disease, are costly. Therefore, identifying potentially inappropriate or unnecessary prescribing can help reduce the burden of healthcare costs related to dialysis for both patients and health insurers, including the MOHSS. A study conducted in a dialysis facility in Detroit (US) reported that the inclusion of a clinical pharmacist resulted in a total cost saving of almost US$500,000. ²⁷ In most dialysis centres in Namibia, pharmacists are not part of the multidisciplinary team responsible for direct care of dialysis patients. This study was necessary to identify gaps in the pharmaceutical care of haemodialysis patients that pharmacists in Namibia can fill.

Aim

The main purpose of this study was to assess the appropriateness of medication therapy among patients who underwent maintenance haemodialysis at two dialysis centres in Windhoek, Namibia. Further, the study also compared prescribing patterns between state-funded and privately-funded haemodialysis patients.

Ethics Approval

Ethical approval was obtained from the University of Namibia Human Research and Ethics Committee (Ref: SOP0001) and the MOHSS (Ref: 17/3/3/MS). Written permission for data collection was also obtained from the management of the dialysis centres. No identifiable patient information was collected for the purpose of this study.

Methods

Study Design, Setting, and Participants

A retrospective, quantitative, analytic design involving review of clinical records of patients attending haemodialysis at two dialysis centres in Windhoek, Namibia, was conducted from July 2022 to December 2022. These two outpatient dialysis centres are private health facilities; however, they attend to both state and privately funded patients. In this study, state patients were individuals who typically received healthcare services at state-run health facilities but received renal care at the dialysis centres and their renal care costs were covered by the MOHSS. Privately funded patients were those patients whose renal care costs were covered by medical insurance, or they paid out of pocket.

Sample Size and Eligibility Criteria

The minimum sample size for this study was determined based on the prevalence of inappropriate medication selection and inappropriate dosage adjustment among CKD patients reported in studies conducted in African countries, which ranged from 27%-43% and 14-35%, respectively.^18-21 Using Epi Info version 7.2.4.0 (CDC, Atlanta, GA, USA), a single-population proportion formula was used. Assuming a prevalence of inappropriate drug selection of 43% ²¹ and a prevalence of inappropriate dosage adjustment of 35% ²⁰ among dialysis patients in Windhoek, with a 95% confidence interval, 80% power, the minimum sample sizes were determined to be 102 and 100 for these two main outcome measures, respectively. At the time of the study, these two dialysis centres each attended to an average of 70 patients per month.

All patients who were diagnosed with ESRD regardless of age and gender, underwent haemodialysis sessions for at least 1 month, had comorbid conditions managed with at least one medication and attended haemodialysis sessions at one of the two centres between July 2022 and December 2022 were included in the study. Patients whose clinical records had missing main outcome measures were excluded. However, when a patient had missing information for variables other than the main outcome measures, such patients were included in the study but excluded from the analysis for that particular variable. For example, the computed median body mass index excluded patients whose recent weight or height were missing. In cases where a diagnosis was not recorded, and there was no clinical suspicion of such a diagnosis based on laboratory results, vital signs, or medications prescribed, it was assumed that that the condition was not present. The same was assumption was applied for missing laboratory results that were either not recorded or whose printed results were not attached to the patient’s clinical records. This approach is consistent with standard clinical practice.

There were notable differences in the use of laboratory tests used to screen for acid-base disorders, between state-funded and privately-funded patients. This indicated variations in monitoring practices. Since the analysis was based on the data available, these differences could have influenced the variations reported in the types of comorbidities and thereby affecting comparability between the two groups. This was recognised as a possible confounding factor during interpretation of results, however, since this differences in monitoring practices was not measured, it could not be controlled for.

Data Collection Procedure and Study Outcomes

Data for all patients that met the inclusion criteria was collected using a pretested data collection tool (Supplementary Material 1). Information on the types of medicines prescribed was obtained from the clinical records and were later classified according to the Anatomical Therapeutic Chemical (ATC) classification. ²⁸ Medications containing combinations were considered as single medicine, represented by one ATC code.

For each patient, each of their prescribed medicines were assessed to determine whether they required dosage adjustment in haemodialysis based on the recommendations from the Drug Prescribing in Renal Failure Handbook. ²⁹ The Oxford Handbook of Dialysis, ³⁰ British National Formulary (BNF), ³¹ South African Medicines Formulary (SAMF), ³² and Summary of Product Characteristics ³³ were also used as secondary sources of information on renal dosage adjustment. For this study, a dosage adjustment error was defined as failure to change a dose or dosing interval to that recommended for patients on haemodialysis. A drug dosage was rated as ‘inappropriate’ if the dosage prescribed was not in conformity with the recommended adjustment in haemodialysis. Fixed dose combinations were rated as ‘inappropriate’ if any of the active pharmaceutical ingredients in the combination was inappropriately dosed.

Each prescribed medicine was also assessed to determine appropriateness of medication choice in managing comorbid conditions. The primary references used to check for the appropriateness of medication choice was the KDIGO Clinical Practice Guidelines.^34-39 Other sources such as the British National Formulary (BNF), South African Medicines Formulary (SAMF), Oxford Handbook of Dialysis were all used as secondary sources in case of missing, unclear, or conflicting information in the primary reference source.^30-32

Each prescribed medicine was studied to determine the appropriateness of medication choice by checking several aspects; they were assessed on whether they were the correct drug to manage the comorbid condition(s), any contraindications in ESRD, use of medicine without indication, presence of clinically significant major drug interactions between the medicines used, therapeutic duplications as well as any drug allergies the patient may have had. Medication choices that met the above criteria were rated as ‘appropriate’ and those that did not meet one or more of the criteria were rated as ‘inappropriate’.

Statistical Analysis

Data were entered in Microsoft Excel and exported to IBM SPSS 26 for analysis. ⁴⁰ Descriptive statistics were used to summarise the participants’ demographics. To compare the demographics and prescribing patterns between state- and private-funded patients, Chi-square or a Fisher’s exact tests were used for categorical variables, and unpaired T-test or Mann-Whitney U test was used for continuous and discrete variables. A P-value < .05 was considered statistically significant.

Results

Patients’ Socio-Demographic and Clinical Characteristics

A total of 181 clinical records were reviewed and 147 were included in the study. The 34 clinical records excluded from the study belonged to patients who either had acute kidney injury or were temporary visitors from other parts of the country and were missing information on the main outcome measures.

Of the included records, slightly more than half (n = 78, 53.1%) were state funded. Of the state patients, 53.8% were female compared to 37.7% of private patients (P = .05). State funded patients tended to be younger (42.0 ± 13.3 years vs 53.2 ± 14.9 years; P < .001). Private patients had more comorbidities (4 ± 3; P = .002) and a higher average number of medicines per patient (12.4 ± 3.9; P = .004) compared to state patients. The most common co-morbid conditions in both state and private patients were hypertension (n = 74, 98.6% vs n = 69, 100%) and anaemia (n = 72, 92.5% vs n = 64, 92.8%). Details of the demographic differences are shown in Table 1.

Table 1.

Demographics and Clinical Data of Patients on Maintenance Haemodialysis (n = 147).

Characteristics	Total (n = 147)	State (n = 78)	Private (n = 69)	P-value
Gender (n, %)
Male	79 (53.7)	36 (46.2)	43 (62.3)	.05 a
Female	68 (46.3)	42 (53.8)	26 (37.7)
Age, years (mean ± SD)	47.3 ± 15.1	42.0 ± 13.3	53.2 ± 14.9	<.001 b
Body Mass Index kg/m2 (median ± IQR) 1	22.1 (18.8-27)	21.8 (18.6-21.8)	22.4 (19-29.1)	.163 c
Number of comorbidities (median ± IQR)	4 ± 3.9	3 ± 2	4 ± 3	.002 c
Number of medicines per patient (mean ± SD)	11.5 ± 3.5	10.7 ± 2.9	12.4 ± 3.9	.004 b
Major co-morbid conditions
Hypertension	143 (98.6)	74 (94.9)	69 (100)	.175 a
Anaemia	136 (92.5)	72 (92.3)	64 (92.8)	.918 a
Secondary hyperparathyroidism	91 (61.9)	57 (73.1)	34 (49.3)	.003 a
Acid-base disorders	69 (46.9)	6 (7.7)	63 (91.3)	<.001 a
Other cardiovascular diseases #	44 (29.9)	13 (16.7)	31 (44.9)	<.001 a
Hyperkalaemia	40 (27.2)	19 (24.4)	21 (30.4)	.409 a
Diabetes mellitus	32 (21.8)	8 (10.3)	24 (34.8)	<.001 a
Venous thromboembolism	26 (17.7)	12 (15.4)	14 (20.3)	.437 a
HIV infection	23 (15.6)	11 (14.1)	12 (17.4)	.584 a
Genito-urinary disorders	23 (15.6)	9 (11.5)	14 (20.3)	.145 a

^aChi-square test.

^bT-test (independent samples).

^cMann-Whitney U test.

^#Other cardiovascular disorders include; heart failure, rheumatic heart disease, ischaemic heart disease).

¹Data available for 89 patients (State = 51, Private = 38).

Appropriateness of Renal Drug Dosage and Medication Choice

During the study period, 163 unique medicines were prescribed to the 147 participants, classified based on their ATC classification. Out of the 1691 total medicines prescribed, 15.6% (n = 263) required renal dosage adjustment, slightly under half (n = 128, 48.7%) of those were appropriately adjusted (Table 2). The most common inappropriately dosed medicine was perindopril. Of the patients prescribed perindopril, 55 patients were receiving higher than the maximum dose recommended for patients on dialysis. Most of the patients (82%) had at least one inappropriately renally adjusted medicine (Table 2).

Table 2.

Renal Drug Dosage Adjustment Among Patients Undergoing Maintenance Haemodialysis (n = 147).

Variable	n (%)
Number of unique medicines prescribed	163
Medicines with guidelines for dosage adjustment	41 (25.2)
Total medicines prescribed to all patients	1691
Medicines requiring dosage adjustment	263 (15.6)
Medicines appropriately adjusted	128 (48.7)
Medicines inappropriately adjusted	135 (51.3)
Type of dosing error
Underdosage	21 (15.6)
Overdosage	111 (82.2)
Treatment duration too long	3 (2.2)

Among 147 patients included in the current study, 33.3% (n = 49) had one or more inappropriately selected medicines. The most common medication selection errors were identified in the management of CKD complications. In the management of anaemia, 46 patients had an elevated haemoglobin (>10 g/L) and continued to receive erythropoietin. Additionally, 30 patients with elevated ferritin levels were still receiving IV iron sucrose. Eighteen patients were receiving an ACE-I or ARB while having hyperkalaemia (>5.5).

Prescribing Patterns Among State Funded and Privately Funded Patients

There were differences in medication prescribing between state- and privately-funded patients. More private patients were prescribed vitamin D or its analogues (P < .001), phosphate binders (P < .001), antithrombotic agents (P < .001), lipid modifying agents (P < .001), and angiotensin receptor blockers (P < .001). In comparison, state patients were more likely to be prescribed calcium supplementation (P < .001), iron supplementation (P = .025), calcium channel blockers (P < .001), and angiotensin converting enzyme inhibitors (P < .001). There were no differences in the prescribing of erythropoietin, folic acid, other antihypertensives, and anti-infectives. Table 3 shows the medicines prescribed from each of the most prescribed drug classes.

Table 3.

The Most Frequently Prescribed Medicines for Patients Receiving Renal Replacement Therapy.

Medication name	Total (n = 167 (%))	State (n = 78)	Private (n = 69)	P-value
Alimentary tract and metabolism
Vitamin D and analogues	126 (85.7)	59 (75.6)	67 (97.1)	<.001
Calcium supplements (including Calcium gluconate)	88 (59.9)	60 (76.9)	28 (40.6)	<.001
Phosphate binders	48 (32.7)	15 (19.2)	33 (47.8)	<.001
Proton pump inhibitors	41 (27.9)	17 (21.8)	24 (34.8)	.080
Blood and blood forming organs
Antithrombotic agents	107 (72.8)	29 (37.2)	47 (68.1)	<.001
Iron preparations	132 (89.8)	68 (87.2)	50 (72.5)	.025
Folic acid	31 (21.1)	13 (16.7)	19 (27.5)	.111
Erythropoietin	132 (89.8)	67 (85.9)	65 (94.2)	.097
Cardiovascular system
Calcium channel blockers (any)	78 (53.1)	59 (75.6)	19 (27.5)	<.001
Lipid modifying agents	25 (17.0)	3 (3.8)	22 (31.9)	<.001
Beta blocking agents	90 (61.2)	49 (62.8)	53 (76.8)	.066
Alpha-adrenoreceptor antagonists	77 (52.4)	46 (59.0)	31 (45.0)	.089
Diuretics	121 (82.3)	65 (83.3)	56 (81.2)	.730
Angiotensin-converting enzyme inhibitor (ACE-I)	65 (44.2)	46 (59.0)	19 (27.5)	<.001
Angiotensin II receptor blocker (ARB)	53 (32.7)	16 (20.5)	37 (53.6)	<.001
Anti-infective for systemic use
Beta lactam antibiotics	6 (4.1)	2 (2.6)	4 (5.8)	.323
Sulphonamides and trimethoprim	3 (2.0)	1 (1.3)	2 (2.9)	.489
Fluoroquinolone antibiotics	2 (1.4)	1 (1.3)	1 (1.4)	.930
Carbapenems	2 (1.4)	2 (2.6)	0 (0)	.180
Aminoglycosides	2 (1.4)	2 (2.6)	0 (0)	.180
Antiretrovirals for treating HIV	23 (15.6)	11 (14.1)	12 (17.4)	.584

Discussion

Principal Findings

This study found that fewer than half of medicines requiring dosage adjustment were appropriately adjusted. One-third of patients had at least one or more inappropriately selected medicines. Further, there were some differences in the types of medicines prescribed to state- and privately-funded patients.

Demographics and Clinical Data of Patients on Maintenance Haemodialysis

Interestingly, state-funded patients had a lower prevalence of comorbidities compared to privately funded patients. This may be explained by the fact that state-funded patients were on average 11 years younger compared to privately funded patients. However, it may also be influenced by limitations in screening for complications. For example, for state patients, urea and electrolyte (U&E) screening usually only includes sodium, potassium, urea, and serum creatinine. More patients in private were monitored with a basic metabolic panel (BMP) which includes bicarbonate and chloride in addition to the others in the U&E. Therefore, it is not surprising that patients in state facilities had fewer documented acid-base disorders, as this is usually identified with the serum bicarbonate which was not available for most state-funded patients.

Uncontrolled hypertension was the most common risk factor for CKD in both state and private patients. A diagnosis of hypertension at a younger age has been significantly associated with higher risk of CKD. ⁴¹ Two studies conducted in South Africa and Kenya also reported an mean age of patients with CKD similar to that of state funded patients in our study.^18,42 These findings show that young people of working age are being affected by kidney disease across sub-Saharan Africa, highlighting opportunities for intervention to prevent or delay renal replacement therapy.

Appropriateness of Medication Dosage and Selection

Only 18% of the 147 patients had all their medicines appropriately adjusted. Out of the total medicines prescribed to all patients, 263 (15.6%) medicines required dosage adjustment and slightly under half (n = 128, 48.7%) were appropriately adjusted. With overdosage being the most common category of drug dosage adjustment errors. This highlights the opportunity for improvement in medication dosing and pharmacists could play a key role. These findings are similar to those reported by studies undertaken in other LMICs such as Pakistan^43,44 and Palestine. ⁴⁵ These countries have reported that dosage adjustment errors are one of the most prevalent drug related errors in CKD patients. The high prevalence of dosage adjustment errors in LMICs in Asia^43-46 and Africa^18,47 is not unexpected as dose adjustment in patients with renal impairment is rarely done in most settings. The failure to adjust doses has been attributed to prescribers underestimating the potential adverse outcomes of lack of dose adjustment. ⁴⁵ In contrast, a study carried out in the US, reported a much lower prevalence of dosage adjustment errors in haemodialysis patients. ²⁷ There are a variety of potential factors leading to fewer errors. For instance, CKD patients in high income countries are more likely to be attended to by a multidisciplinary health care team comprising doctors and clinical pharmacists. Although the practice of pharmacy has transformed in many parts of the world to become more patient-oriented, in some LMICs, pharmacists still have more traditional roles focused on pharmaceutical products and dispensing. ⁴⁸ These pharmacists rarely have access to important clinical information such as renal function test results which would be necessary to make appropriate renal dosing recommendations. In high-income countries, electronic medical records can allow team members real-time access to all the clinical data for a patient, particularly in hospital settings. These systems also have alerts that may make recommendations on renal dosing with medication order entry.

Medications were appropriately selected in 80.3% (n = 118) of the patients, while the remaining 19.7% had at least one or more inappropriately selected drugs. Although patients were managed with a relatively high number of medicines (on average, 12 per patient), the findings on medication selection appropriateness suggest that it was clinically appropriate in most cases. In recent years some studies have suggested redefining polypharmacy from merely counting number of medicines prescribed to adopting terms such as ‘appropriate or inappropriate polypharmacy. ⁴⁹ The most common medication selection errors were identified in the management of CKD complications. Other studies have reported a much higher prevalence of medication choice errors ranging from 27.7% to 60.3%.^19,50,51 In the current study, patients were normally seen by one prescriber at the dialysis centre who was responsible for initiating therapy as well as changing dialysis regimens; having consistent providers has been shown to be a factor that may lead to less medication choice errors. ⁵² Previous studies have demonstrated the role of pharmacists in identifying and intervening in medication dosing and selection errors when they are directly involved in the care of haemodialysis patients.^27,53-55 This study suggests that pharmacists in Namibia may have an important role in improving medication selection and dosage adjustment in patients on haemodialysis.

Comparison of Prescribing Patterns Between State- and Privately-Funded Patients

More private patients were prescribed vitamin D or its analogues (97%.1%) and phosphate binders (47.8%) (P < .001). This highlights a possible overuse of these agents in private patients since less than half of private patients had secondary hyperparathyroidism and active vitamin D levels were rarely monitored. State patients were less likely to receive phosphate binders (P < .001) even though more state patients compared to private patients had secondary hyperparathyroidism (73.1%). This may be because phosphate binders are not part of the Namibia Essential Medicines List (NEMLIST) ⁵⁶ due to their high cost and are therefore not readily available or routinely prescribed in state hospitals. Private patients were also more likely to be prescribed antithrombotic agents (P < .001) and lipid modifying agents (P < .001). This may be due to higher prevalence of cardiovascular diseases (P < .001) and diabetes (P < .001) among private patients, or less screening for cardiovascular disease risk for state patients.

State patients were more likely to be prescribed calcium supplementation (P < .001), mainly calcium gluconate. The frequent prescribing of calcium gluconate in state patients’ needs to be explored further. It is not known whether it was used in place of phosphate binders or as a calcium supplementation. Calcium gluconate 300 mg tablets are the only oral calcium supplements included on the NEMLIST. This tablets contain approximately 28 g elemental calcium and the commonly prescribed dose of 600 mg thrice daily is not sufficient to manage hypocalcaemia in CKD. ⁵⁷ There is limited evidence to support the use of calcium gluconate as a phosphate binder hence it is not recommended in international guidelines.^36,58

Regarding hypertension management, state patients were more likely to be prescribed calcium channel blockers (P < .001), and angiotensin converting enzyme inhibitors (P < .001). In state hospitals in Namibia, angiotensin converting enzyme inhibitors are preferred over angiotensin receptor blockers and are prescribed more frequently for management of hypertension even in patients without CKD. ⁵⁹ In comparison, private patients were more likely to be prescribed an angiotensin receptor blockers (P < .001), which may be due to prescribing preferences and better access to this class of medicines.

There were several limitations in our study. Firstly, this study made use of data from manual patient clinical records, for this reason some data may have been recorded wrongly or omitted. The exclusion of patients with missing main outcome measures may have introduced selection bias. Nonetheless, all patient records that were reviewed and met the inclusion criteria were included in the study, resulting in a more representative sample. Differences in monitoring between state- and privately-funded patients may have influenced types of comorbidities identified and prescribing patterns observed between the two groups. For example, privately-funded patients with more access to laboratory monitoring are more likely to be diagnosed with a complication of CKD and are more likely to have their treatment adjusted accordingly. Actual utilization of these prescribed medicines by patients could not be confirmed and information on other medicines that the patients may have been taking such as self-medications or herbal medicines was not obtained either. Interviewing patients may have also revealed other issues such as poor adherence, inappropriate timing of drug administration, and may have helped identify drugs that may no longer be necessary for the patient.

Impact of the Study

The findings of our study support the need for a more multidisciplinary approach involving nephrologists, general practitioners, nurses, laboratory personnel and pharmacists in the care of patients receiving dialysis, to improve rational prescribing practices. Continuing professional development activities designed specifically for pharmacists caring for patients on dialysis are needed to make them a proactive and effective member of the team. Currently, there is only a brief section in the NSTG addressing the diagnosis of renal diseases, and haemodialysis is simply mentioned as an option for patients with renal disease. ⁵⁹ However, the growing prevalence of CKD in Namibia calls for the inclusion of more evidence-based information on the management of common complications and comorbidities in patients receiving haemodialysis, as well as recommendations for appropriate clinical monitoring to the next edition of the NSTG. This change would drive the needed revision of the NEMLIST to include more medicines for non-communicable diseases such as CKD and its complications, ensuring alignment between the two documents. The NEMLIST has historically prioritised communicable diseases, however, with shifting disease demographics in Namibia, there is a need for equitable health resource allocation and improved access to medicines used in the management of non-communicable diseases.

Conclusion

Patients on maintenance haemodialysis were managed with a relatively large number of medicines although the medicines are appropriately selected for their indications in most cases, renal dosage adjustment was inappropriately undertaken in most cases. The differences in the prescribing patterns of vital medicines such as phosphate binders, between state funded- and privately-funded patients warrants further investigation. A wide variety of medicines were prescribed to all patients. There is still much work needed to optimise drug therapy, improve patient outcomes and ensure safe management of these patients. Clinical pharmacy services would therefore be of great value in the care of dialysis patients in Namibia.