The search yielded a total of 5,487 published articles. Eighty-five duplicates were identified and removed. Of the remaining 5,402 articles, a total of 12 articles that were editorials, review articles, systematic reviews, study protocols and a study published in French were excluded leaving 5,390 articles whose titles and abstracts were screened.
Of these, 5,329 articles were excluded leaving 61 articles whose full texts were extracted and assessed for eligibility. Only 21 original research articles met the inclusion criteria described above and were included in the final systematic review [11–31] (Fig. 1).
Study Characteristics, Methodological Quality And Risk Of Bias
All selected studies were cross-sectional in design. There was an observed heterogeneity of the study findings. Among the 19 studies conducted in one region of SSA, the majority were from the Eastern region (n = 11, 58%) [11–21]. About 26% (n = 5) [22–26]and 16% (n = 3) [27–29] of the remaining studies were conducted in the Southern and Western region respectively. Two studies were conducted in more than one region of SSA [30, 31]. According to the quality assessment tool for observational cohort and cross sectional studies, all the original cross-sectional studies included in the systematic review were considered as having a low risk of bias (Tables 4). The majority of the studies had low methodological quality (Table 3).
Availability Of Essential Medicines For Dm
The availability of the different essential medicines for DM as recommended by the WHO PEN are shown in Table 2.
Table 2
Availability and affordability of essential medicines and diagnostic tests for diabetes in Sub-Saharan Africa.
Study, year and reference | Country (ies) where study was done and study period | No. of health facilities surveyed | Number of essential medicines and diagnostic tests studied | Key study findings |
A: Eastern region | | | | |
1. Kibirige D et al, 201710 | -Uganda. -15th January 2017 to 28th February 2017 | -22 public hospitals, 23 private hospitals and 100 privately owned pharmacies. | - Insulin (Short-acting, Intermediate-acting and Pre-mixed). -OHA (Glibenclamide, Glimepiride and Metformin) - ACEI (Captopril) -Aspirin -Statins (Simvastatin, Atorvastatin and Rosuvastatin) Diagnostic tests -Glucometers -Serum creatinine -Lipid profile -Microalbuminuria tests. -Serum troponin -Urine protein and ketone testing strips -ECG -HbA1c -Serum ketone tests | Availability of medicines - Intermediate-acting insulin-34.7% - Pre-mixed insulin-60.1% - Short-acting insulin-68.8% - Statins-84% - Aspirin-95.1% - ACEI-96.5% - Any OHA present-100%. Availability of diagnostic tests - Microalbuminuria tests-6.8% - Serum ketones-11.4%. - Serum troponin-43.2%. - HbA1c tests-43.2% - ECG-54.6% - Lipid profile-65.9% - Serum creatinine-86.4% - Glucometers-97.7% - Urine protein and ketone testing strips-100% Affordability of essential medicines - Short-acting insulin-4.7 days’ wages -Intermediate-acting insulin-4.9 days’ wages -Pre-mixed insulin-4.9 days’ wages. -Metformin 500 mg-2.8 days’ wages. -Glibenclamide 5 mg-0.7 days’ wages. -Glimepiride 2 mg-3.2 days’ wages. -Aspirin-0.9 days’ wages. -Captopril-2.8 days’ wages. -Simvastatin-6.5 days’ wages. -Atorvastatin 20 mg- 7.6 days’ wages. -Rosuvastatin 10 mg- 7.6 days’ wages. Affordability of diagnostic tests - Blood glucose testing-1.1 days’ wages. - Urine protein and ketone testing − 1.3 days’ wages. - Serum ketones testing-2.1 days’ wages. - Serum creatinine-2.4 days’ wages. - Lipid profile-7.5 days’ wages - HbA1c-8.6 days’ wages. - Microalbuminuria-9.6 days’ wages. - ECG-10.7 days’ wages. - Serum troponin- 11.3 days’ wages. |
2. Musinguzi G et al, 201511. | -Uganda. -June to October 2012. | -126 health facilities (74 public and 52 private) | -ACEI. | Availability of any of the essential medicines -ACEIs: 22.2%. |
3. Bekele A et al, 201712. | -Ethiopia. -March to July 2014. | -873 health facilities (51%-public and 45%-private facilities). | -Insulin (type not specified). -Glibenclamide. -Metformin. -Glucometers. -Urine protein and ketone testing strips. | Availability of essential medicines -Insulin: 9%. -Metformin: 11%. -Glibenclamide: 28%. -Availability of diagnostic tests -Glucometers: 40%. -Urine protein and ketone testing strips : 56%. |
4. Carlson S et al, 201713 | -Uganda and Kenya. -2011 and 2012. | -340 health facilities (58.2% were public). | -ACEI (Captopril or lisinopril). -ECG. | Availability of essential medicines -ACEIs (captopril or lisinopril): 45.6%. Availability of diagnostic tests. -ECG: 9.1%. |
5. Armstrong-Hough M et al, 201814. | -Uganda. -2013 | -196 health facilities (63.8% were public). | - Short-acting insulin. -Metformin. -Glibenclamide -Statins (simvastatin). -Any ACEI | Availability of essential medicines DM - Short-acting insulin: 11.2%. -Metformin: 23.5%. -Glibenclamide: 25.5% -Statins (simvastatin): 3.1%. -Any ACEI: 16.8%. |
6. Bintabara D et al, 2018. | -Tanzania. -2014-2015. | -725 health facilities (68% were public) | -ACEI. | Availability of essential medicines -Any ACEI: 21%. |
7. Katende D et al, 2015. | -Uganda. -November 2012- April 2013 | -28 health facilities (86% were public). | -Metformin. -Glucometers | Availability of essential medicines -Metformin: 17.9%. Availability of diagnostic test -Glucometers: 32%. |
8. Peck R et al, 2014 | -Tanzania. -November 2012- May 2013. | -24 health facilities (75% were public). | -Metformin. -Glucometers. | Availability of essential medicines -Metformin: 33.3%. Availability of diagnostic test -Glucometers: 33.3%. |
9. Getachew T et al, 2017 | -Ethiopia -2016 | -547 health facilities (59% were public) | -Glicazide or Glipizide. -Insulin (any type). -Glibenclamide. -Metformin. -ACEI. -Aspirin. -Glucometers. -Urine protein and ketone testing strips. | Availability of essential medicines -Glicazide or Glipizide: 4%. -Insulin (any type): 18%. -Glibenclamide and metformin: 31%. -ACEI: 25%. -Aspirin: 53%. Availability of diagnostic tests -Glucometers: 66%. -Urine protein and ketone testing strips: 89%. |
10. Rogers HE et al, 2018 | -Uganda. -2013. | -53 public health facilities. | -Metformin. -Any sulphonylurea. -Insulin (Ultra short- acting, short-acting, intermediate-acting and long-acting). -Any ACEI. -Glucometers. -ECG. -Serum creatinine. -Lipid profile. -Urine protein and ketone testing strips. - HbA1c. - Microalbuminuria. | Availability of essential medicines -Short-acting insulin: 52.8% -Intermediate-acting insulin: 47.2%. -Long-acting insulin: 52.8%. -Any sulphonylurea: 81.1%. -Metformin: 92.5%. -ACEI: 67.9%. Availability of diagnostic tests -ECG: 5.7%. -HbA1c: 9.4%. -Microalbuminuria: 13.2%. -Lipid profile: 28.3%. - Serum creatinine: 43.4%. - Glucometers: 62.3%. -Urine protein and ketone testing strips: 92.5%. |
11. Whyte SR et al, 2015. | -Uganda. -November 2011 and February 2012 | -6 health facilities (83.3% were public). | - Short-acting insulin. -Intermediate-acting insulin. -Pre-mixed insulin. -Glibenclamide. -Metformin. -ACEI (Captopril). | Availability of essential medicines -Glibenclamide: 50%. -Metformin: 16.7%. -Either short-acting, intermediate-acting or pre-mixed insulin: 16.7%. -ACEI: 50%. Availability of diagnostic test -Glucometers: 33.3%. |
B: Southern region | | | | |
12.Beran D et al, 2005 | - Mozambique and Zambia. - April and May 2003 in Mozambique and Sept and October 2003 in Zambia. | -5 hospitals and 6 lower tier health centres in Mozambique. -13 hospitals and unspecified number of referral health centres in Zambia. | -Insulin (pre-mixed, short-acting and intermediate-acting insulin). -Glucometers. | Availability of essential medicine: -Insulin: 0% in 6 surveyed health centres and 20% in 5 surveyed hospitals in Mozambique. -42% in the surveyed referral health centres and 100% in all 13 surveyed hospitals in Zambia. Availability of diagnostic test: -Glucometers: 6% in Mozambique and 25% in Zambia. |
13. Mendis S et al, 2007 | -Malawi − 2005 | -20 public and 16 private health sector facilities. | - Insulin (3 types) - Glibenclamide - Metformin - ACEI (captopril and enalapril). - Lovastatin (statin) | Availability of the essential medicines in public and private sector respectively: - Short-acting insulin: 25% and 6%. -Insulin zinc suspension: 30% and 25%. -Insulin isophane: Both 0%. -Captopril (An ACEI): 5% and 63%. -Enalapril (An ACEI): 0% and 81%. -Lovastatin (Statin): 0% and 56%. Affordability of the essential medicines. -Monotherapy with LPG Glibenclamide and Metformin also cost < 1 days’ wages. -40 IU of intermediate insulin (innovator brand): 19.6 days’ wages. -Monotherapy with LPG oral captopril and enalapril (ACEI or ARB) cost < 1 days’ wages. |
14. Kalungia CA et al, 2017. | -Zambia. -January to June 2016. | -15 public health facilities. | -Insulin (short-acting and long-acting). -Glibenclamide. -Metformin | Availability of the essential medicines -Short-acting insulin: 22.2%. -Long-acting or intermediate-acting insulin: 37.8%. -Metformin: 51.1%. -Glibenclamide: 51.1%. |
15. Mhlanga B et al, 201424 | -Swaziland -December 2012-January 2013 | -10 public facilities and 10 private retail pharmacies. | -Glibenclamide. -Metformin. -ACEI (Captopril and Enalapril) | Availability of the essential medicines -Glibenclamide: 90%. -Metformin: 100%. -Captopril: 90%. Affordability of the essential medicines -The lowest priced generic Metformin, Glibenclamide and Captopril: 1.2 days’ wages. |
16. Chikowe I et al, 2018. | -Malawi. -November to December 2016. | -55 health facilities (76.4%-public facilities). | -Insulin (type not specified). -Glibenclamide. -Metformin. -Glucometers. | Availability of the essential medicines -Insulin: 1.8%. -Glibenclamide: 9.1%. -Metformin: 14.5%. Availability of the diagnostic tests -Glucometers: 38.2%. |
C: Western region | | | | |
17. Jingi A et al, 2014 | -Western Cameroon. -2012 | -2 private and 9 public health facilities. | - Insulin (Short-acting, Intermediate-acting and Pre-mixed). -OHA (Glibenclamide and Metformin) - ACEI (Captopril and Ramipril) -Statins (Simvastatin) -Aspirin Diagnostic tests -Glucometers -HbA1c -Serum creatinine. -Lipid profile -Urine protein and ketone testing strips. -ECG | Availability of medicines: -Intermediate-acting insulin (Insulatard): 10% -Glibenclamide, Metformin, Short-acting insulin (Actrapid) and Pre-mixed insulin (Mixtard): all at 80%. - Simvastatin-10%. - Ramipril-20%. - Captopril-30%. - Aspirin: 70% Availability of diagnostic tests -ECG: 10% -HbA1c: 20% -Lipid profile: 40% -Serum creatinine: 80% -Urine protein and ketone testing strips: 90%. -Glucometers: 100%. Affordability of essential medicines -Glibenclamide 5 mg-0.34 days’ wages. -Metformin 500 mg-0.67 days’ wages. -Short-acting insulin-3.85 days’ wages. -Intermediate-acting insulin-3.85 days’ wages. -Pre-mixed insulin-18.7 days’ wages. -Aspirin-0.03 days’ wages. -Captopril 25 mg-6.41 days’ wages. -Simvastatin 20 mg- 30.51 days’ wages. Affordability of diagnostic tests -Urinalysis: 1.07 days’ wages. -Glucose testing: 1.34 days’ wages. -Serum creatinine: 3.21 days’ wages. -Lipid profile: 3.59 days’ wages. -ECG: 10.7 days’ wages. -HbA1c test: 12.58 days’ wages. |
18. Nyarko KM et al, 2016 | -Ghana. − 9th June to 28th June 2013 | -24 health facilities (21 public and 3 private hospitals). | -Long-acting insulin. - Short-acting insulin. -Glibenclamide. -Metformin. -ACEI (Enalapril or lisinopril). -Statins. -Aspirin. -Lipid profile. -Serum creatinine. -Glucometers. -Urine protein and ketone testing strips. | Availability of essential medicines -Long-acting insulin: 16.7%. -Short-acting insulin: 20.8%. -Glibenclamide: 20.8%. -Metformin: 25%. -Statins: 12.5%. -Enalapril or lisinopril: 25%. -Aspirin: 79.2%. Availability of diagnostic tests -Serum troponin: 0%. -Lipid profile: 16.7% -Serum creatinine: 16.7%. -Glucometers: 25%. -Urine protein and ketone testing: 33.3%. |
19.Okpetu EI et al, 2018 | -Nigeria. -June to July 2013. | -6 public primary health facilities. | -Glucometers. -Urine protein and ketone testing strips. | Availability of diagnostic tests -Glucometers: 33.3%. -Urine protein and ketone testing strips: 100%. |
D: Multi-regions | | | | |
20. Cameron A et al, 2009 | -36 LMIC (11 African countries included). -Western region (Cameroon, Chad, Ghana, Mali and Nigeria), Eastern region (Ethiopia, Kenya, Sudan, Tanzania and Uganda) & Southern region (South Africa). -May 2008 | -45 national and subnational surveys. | -Glibenclamide. | -Mean availability of glibenclamide in 8 African countries reported: 37.3% and 60.6% in the public and private sector respectively. -Affordability of LPG glibenclamide in 7 African countries was 1.1 and 1.8 mean days’ wages in the public and private sector respectively. |
21. Mendis S et al, 201232 | -8 LMIC (3 African countries) - Western region (Benin) & Eastern region (Eritrea and Sudan). -January 2009 to January 2011. | − 30 health facilities in the 3 African countries. | -Long-acting insulin. -Short-acting insulin. -Metformin. -Glibenclamide. -Enalapril. -Simvastatin or Lovastatin. -Aspirin. -Urine protein and ketone testing strips. -Glucometers. -Lipid profile. -Serum creatinine. | Availability of essential medicines in Benin, Eritrea and Sudan respectively: -Long-acting insulin: 0%, 0% and 21.4%. - Short-acting insulin: 0%, 0% and 28.6%. -Metformin: 25%, 0% and 42.9%. -Glibenclamide: 41.7%, 0% and 71.4%. -Enalapril: 33.3%, 0% and 28.6%. -Simvastatin or Lovastatin: 8.3%, 0% and 35.7%. -Aspirin: 100% in all 3 countries. Availability of diagnostic tests in Benin, Eritrea and Sudan respectively: -Urine protein and ketone testing strips: 100%, 67% and 92%. -Glucometers: 67%, 17% and 75%. -Lipid profile: 25%, 0% and 33%. -Serum creatinine: 33%, 0% and 58%. -Serum troponin: 8%, 0% and 8%. |
Insulin
Generally, availability of insulin as reported by the majority of the studies was sub-optimal basing on the recommended WHO GAP goal of ≥ 80%. Availability of insulin of any type ranged from 0% in Mozambique [22], Benin and Eritrea [31] to 100% in Zambia [22].
In the surveyed health facilities, availability of short-acting insulin was 0% in Benin and Eritrea [31], 6% in Malawi (public health facilities) [23], 11.2% in one study in Uganda [15], 20.8% in Ghana [28], 22.2% in Zambia [24], 25% in Malawi (private health facilities) [23], 28.6% in Sudan [31], 52.8% and 60.1% in two other studies in Uganda [11, 20] and 80% in Cameroon [27].
Availability of intermediate-acting insulin was 0% in Malawi [23], 10% in Cameroon [27] and 16.7%, 34.7% and 47.2% in three studies conducted in Uganda [11, 20, 21].
Availability of pre-mixed insulin was reported by only two studies conducted in Uganda (16.7% and 60.1%) [11, 21] and in Cameroon (80%) [27].
Oral hypoglycaemic agents (OHA)
Availability of oral hypoglycaemic agents (OHA) in the surveyed health facilities ranged from 0% in Eritrea (metformin and glibenclamide) [31] to 100% in Uganda (metformin and glibenclamide/glimepiride) [11] and Swaziland (metformin) [25]. The most studied OHA were glibenclamide and metformin. In addition to those two drugs, one study from Ethiopia also assessed the availability of gliclazide or glipizide which was reported to be very low (4%) [19].
The documented availability of metformin was 11% and 31% in Ethiopia [13, 19], 14.5% in Malawi [26], 25% in Ghana and Benin [28, 31], 33.3% in Tanzania [18], 42.9% in Sudan [31], 51.1% in Zambia [24] and 80% in Cameroon [27]. The additional four studies conducted in Uganda reported the availability of metformin ranging from 16.7–92.5% [15, 17, 20, 21].
The availability of glibenclamide was reported to be 9.1% in Malawi [26], 20.8% in Ghana [28], 25.5%, 50% and 81.1% in three other studies in Uganda [15, 20, 21], 28% and 31% in two studies in Ethiopia [13, 19], 41.7% in Benin [31], 51.1% in Zambia [24], 71.4% in Sudan [31], 80% in Cameroon [27] and 90% in Swaziland [25].
Angiotensin Converting Enzyme Inhibitors (ACEI)
Availability of angiotensin converting enzyme inhibitors (ACEI) ranged from 0% in Eritrea [31] to 96.5% in Uganda [11]. Except for the studies conducted in Swaziland [25], Malawi [23] and two studies conducted in Uganda [11, 20], the remaining nine studies documented low availability of ACEI of < 50% [12, 14–16, 19, 21, 27, 28, 31].
Aspirin
Availability of aspirin was investigated by five studies and was reported present in 53% in Ethiopia [19], 70% in Cameroon [27], 79.2% in Ghana [28], 95.1% in Uganda [11] and 100% in Benin, Eritrea and Sudan [31].
Statins
Five studies reported about the availability of statins ranging from 0% in public facilities in Malawi [23] and Eritrea [31] to 84% in Uganda [11]. The documented availability in other studies was 3.1% in another study conducted in Uganda (simvastatin) [15], 8.3% in Benin (simvastatin) [31], 10% in Cameroon (simvastatin) [27], 35.7% in Sudan (simvastatin) [31] and 56% in private health facilities in Malawi (lovastatin) [23].
Availability Of Diagnostic Tests For Dm
A total of 14 studies (66.7%) investigated the availability of at least one diagnostic test for DM. Among the recommended diagnostic tests for DM in the WHO PEN, availability of blood glucose tests was the most investigated (13 studies, 92.9%) [11, 13, 14, 17–22, 26–29, 31] while microalbuminuria tests were the least investigated (2 studies, 14.3%) [11, 20].
The availability of the diagnostic tests for DM as recommended by the WHO PEN is as shown in Table 2.
Blood glucose tests
Availability of blood glucose tests ranged from 6% in a study conducted in Mozambique [22] to 100% in a study conducted in Cameroon [27]. Only two studies reported optimal levels of availability of blood glucose tests of ≥ 80% [11, 27] while one study conducted in Sudan documented close to optimal levels of availability of 75% [31]. Availability of blood glucose tests in the remaining studies was < 70% [13, 17–22, 26, 28, 29, 31].
Urine protein and ketone tests
Urine protein and ketone tests were the second most investigated diagnostic test (8 studies, 57%) [11, 13, 19, 20, 27–29, 31]. The availability ranged from 33.3% in Ghana [28] to 100% in Uganda [11], Nigeria [29] and Benin [31]. Other studies conducted in Ethiopia [13, 19], Eritrea [31], Cameroon [27], Sudan [31] and Uganda [20] reported availability of 56% and 89%, 67%, 90%, 92% and 92.5% respectively.
Lipid profile tests
Availability of lipid profile tests was reported by five studies and this ranged from 0% in Eritrea [31] to 65.9% in one study conducted in Uganda [11]. The remaining studies reported availability of 16.7% in Ghana [28], 25% in Benin [31], 28.3% in Uganda [20], 33% in Sudan [31] and 40% in Cameroon [27].
Serum creatinine tests
Availability of serum creatinine tests was also reported by five studies [11, 20, 27, 28, 31]. Optimal availability of ≥ 80% was only noted in two studies conducted in Uganda [11] and Cameroon [27].
Electrocardiography (ECG)
The documented availability of ECG as reported by four studies was 5.7% and 54.6% in Uganda [11, 20], 9.1% in a study conducted in Uganda and Kenya [14] and 10% in Cameroon [27].
Microalbuminuria tests
Availability of microalbuminuria tests was reported by only two studies conducted in Uganda both documenting low levels of 6.8% and 13.2% respectively [11, 20].
Glycated haemoglobin (HbA1c) tests
Three studies conducted in Cameroon [27] and Uganda (n = 2) [11, 20] investigated the availability of HbA1c tests, noting their presence in 20%, 43.2% and 9.4% of the health facilities that were surveyed respectively.
Serum troponin tests
Availability of serum troponin tests as reported by three studies was 0% in Ghana [28] and Eritrea [31], 8% in Benin and Sudan [31] and 43.2% in Uganda [11].
Serum ketone tests
Availability of the serum ketone tests was investigated by only one study which reported them available in only 11.4% of the surveyed health facilities, which were all private hospitals [11].
Affordability Of Essential Medicines For DM
Affordability of any of the essential medicines for DM was investigated by five studies [11, 23, 25, 27, 30] (summarised in Table 2).
Oral hypoglycaemic agents
The lowest priced generic (LPG) glibenclamide cost less than 2 days’ wages in all the studies [11, 23, 25, 27, 30]. Except for the study conducted in Uganda [11], LPG metformin cost ≤ 1.2 days’ wages in the rest of the studies [23, 25, 27]. One study assessed the cost of the newer generation sulphonylurea-glimepiride whose monthly dose cost 3.2 days’ wages [11].
Insulin
Affordability of insulin was assessed in three studies [11, 23, 27]. Short-acting and intermediate-acting insulin cost 4.7 and 4.9 days’ wages respectively in Uganda [11]. The cost of short- and intermediate-acting insulin in Cameroon was similar (3.85 days’ wages) [27]. A high cost of the innovator brand of intermediate-acting insulin was reported in Malawi (19.6 days’ wages) [23]. A high cost of pre-mixed insulin was noted in Cameroon (18.7 days’ wages) [27] compared to Uganda (4.9 days’ wages) [11].
Aspirin, ACEI (captopril) and statins (simvastatin/atorvastatin/rosuvastatin)
Affordability of any of the three classes of essential medicines above was investigated by four studies [11, 23, 25, 27]. Only two studies assessed affordability of all the three classes of essential medicines above [11, 27].
Aspirin cost less than a days’ wages in the two studies [11, 27]. The cost of captopril and statins greatly varied among the countries. It cost < 1.3 days’ wages in Malawi [23] and Swaziland [25], 2.8 days’ wages in Uganda [11] and 6.41 days’ wages in Cameroon [27]. Any statin cost > 6 days’ wages in Uganda [11] while simvastatin cost 30 days’ wages in Cameroon [27].
Affordability Of Diagnostic Tests
Affordability of diagnostic tests for DM was assessed by only two studies conducted in Uganda [11] and Cameroon [27]. Blood glucose, urine protein and ketone, serum creatinine tests cost < 3.3 days’ wages in both countries [11, 27]. In comparison, the cost of lipid profile testing in Uganda [11] was twice the cost in Cameroon [27] (7.5 and 3.59 days’ wages respectively).
Both ECG and HbA1c cost > 8 days’ wages in both countries with a higher cost documented in Cameroon. The cost of serum ketone, microalbuminuria and serum troponin tests was investigated by only one study reporting costs of 2.1 days’ wages, 9.6 days’ wages and 11.3 days’ wages respectively [11].