Impact of the COVID-19 pandemic on type 2 diabetes care and factors associated with care disruption in Kenya and Tanzania

Figures & data

Table 1. Socio-demographic characteristics of type 2 diabetes patients who participated in the study in Kenya and Tanzania.

Variable		Kenya (N=500)	Tanzania (N=500)
		n (%)	n (%)
Location*	Urban	380 (76.0%)	300 (60.0%)
	Rural	120 (24.0%)	200 (40.0%)
Sex	Female	330 (66.0%)	336 (67.2%)
	Male	170 (34.0%)	164 (32.8%)
Age in years, Mean (SD)		58.2 (12.6)	56.8 (10.2)
Age group	<40 years	39 (7.8%)	36 (7.2%)
	40–49 years	80 (16.0%)	62 (12.4%)
	50–59 years	135 (27.0%)	165 (33.0%)
	60–69 years	155 (31.0%)	206 (41.2%)
	>70 years	91 (18.2%)	31 (6.2%)
Formal education	None	16 (3.2%)	34 (6.8%)
	Primary	208 (41.6%)	298 (59.6%)
	Secondary	201 (40.2%)	126 (25.2%)
	College/University	75 (15.0%)	42 (8.4%)
Occupation	Formal employment	31 (6.2%)	40 (8.0%)
	Farming (small and large scale)	78 (15.6%)	103 (20.6%)
	Self-employed (small and large business)	162 (32.4%)	190 (38.0%)
	Retired	55 (11.0%)	72 (14.4%)
	Unemployed/Homemaker	174 (34.8%)	95 (19.0%)
Family history of diabetes	No	233 (46.6%)	205 (41.0%)
	Yes	267 (53.4%)	295 (59.0%)
Duration from first diagnosis of type 2 diabetes	≥6 years	309 (61.8%)	334 (66.8%)
	<6 years	191 (38.2%)	166 (33.2%)
Health insurance coverage	Uninsured	163 (32.6%)	302 (60.4%)
	Insured	337 (67.4%)	198 (39.6%)
Presence of co-morbidities	Yes	355 (71.0%)	368 (73.6%)
	No	141 (29.0%)	132 (26.4%)
Economic status (ladder scale [1–10])	Lower SES** [1–5]	447 (89.4%)	374 (74.8%)
	Higher SES** [6–10]	53 (10.6%)	126 (25.2%)

*Kenya, Urban - Nairobi and Kiambu counties, Rural - Nyeri and Vihiga counties; Tanzania, Urban - Dar es Salaam, Rural – Morogoro.

**SES, Socio-economic status.

Table 2. Diabetes care before and during the COVID-19 pandemic in Kenya and Tanzania.

		Kenya			Tanzania
Variable		Pre-COVID-19 N (%)	During the pandemic N (%)	Change % (*p value)	Pre-COVID-19 N (%)	During the pandemic N (%)	Change % (*p value)
Blood glucose testing
Where testing took place	Total number (%) who changed testing place		174 (34.8)			62 (12.4)
	Tested at home	148 (29.6)	100 (20.0)	-9.6 (0.002)	13 (2.6)	36 (7.2)	4.6 (<0.001)
	Tested at health facility	197 (39.4)	381 (96.0)	56.6 (<0.001)	433 (86.6)	405 (81.0)	-5.6 (<0.001)
	Tested both at home and health facility	155 (31.0)	19 (3.8)	-27.2 (<0.001)	54 (10.8)	59 (11.8)	1.0 (0.197)
Frequency of testing	Total number (%) who changed frequency of testing		164 (32.8)			89 (17.8)
	Tested at least once in a day	167 (33.4)	150 (30.0)	-3.4 (0.434)	45 (9.0)	52 (10.4)	1.4 (0.280)
	Tested once or twice a week	232 (46.4)	211 (42.2)	-4.2 (<0.001)	28 (5.6)	37 (7.4)	1.8 (0.129)
	Tested at least once a month	87 (17.4)	98 (19.6)	2.2 (0.001)	316 (63.2)	281 (56.2)	-7.0 (<0.001)
	Tested at least once in 3 months	8 (1.6)	19 (3.8)	1.6 (0.784)	62 (12.4)	76 (15.2)	2.8% (0.031)
Medication prescription, access, and use
Medication Prescription	Number (%) who had a change of prescription		96 (19.20)			63 (12.6)
	Prescribed insulin only	45 (9.0)	59 (11.8)	2.8 (0.002)	46 (9.2)	57 (11.4)	2.2 (0.002)
	Prescribed oral hypoglycaemics only	415 (83.0)	402 (80.4)	-2.6 (<0.001)	450 (90.0)	439 (87.8)	-2.2 (0.002)
	Both insulin and oral hypoglycaemics	40 (8.0)	39 (7.8)	-0.2 (<0.001)	4 (0.8)	4 (0.8)	0
Access to medication	Accessed prescribed medications	452 (93.2)	346 (69.2)	24 (<0.001)	482 (96.4)	467 (93.4)	-3.0 (0.009)
Use of prescribed medication	Used medications as prescribed	483 (96.6)	485 (97.0)	0.4 (0.670)	489 (97.8)	480 (96.0)	-1.8 (0.061)
Access to health facilities
Health facility visits	Visited hospital in the last 3 months	445 (89.0)	373 (74.6)	-14.4 (<0.001)	456 (91.2)	428 (85.6)	-5.6 (0.001)
Care access	Successfully accessed healthcare provider	494 (98.8)	464 (92.8)	-6 (<0.001)	498 (99.6)	493 (98.6)	-1.0 (0.096)
	Accessed in-person consultation	488 (98.8)	447 (96.3)	-2.5 (<0.001)	498 (99.6)	491 (98.2)	-1.4 (0.035)

*p values obtained from comparison of variables using the independent samples T-test.

Figure 1. Proportion of cumulative disruptions of diabetes care disruption due to COVID-19 in Kenya and Tanzania.

Disruptions assessed include change in where testing took place, frequency of testing, cost of testing, medication prescription access, cost of prescribed medication and access to health provider.

Figure 2. Levels of diabetes care disruption due to COVID-19 in Kenya and Tanzania.

Disruptions assessed include change in where testing took place, frequency of testing, cost of testing, medication prescription access, cost of prescribed medication and access to health provider.

Table 3. Predictors of severe disruption of diabetes care in Kenya and Tanzania (bivariate analysis).

		Kenya		Tanzania
Variable	Categories (n = Kenya, Tanzania)	Odds ratio (95%CI)	p value	Odds ratio (95%CI)	p value
Location	Urban (n = 380,300)
	Rural (n = 120,200)	0.40 (0.26, 0.63)	<0.001	0.38 (0.26, 0.54)	<0.001
Sex	Female (330,336)
	Male (n = 170, 164)	1.03 (0.71, 1.49)	0.875	0.88 (0.61, 1.27)	0.496
Age group	<40 (n = 39,36)
	40–49 (n = 80,62)	1.56 (0.73, 3.34)	0.255	0.94 (0.41, 2.14)	0.888
	50–59 (n = 135,165)	0.95 (0.46, 1.93)	0.881	0.64 (0.31, 1.33)	0.231
	60–69 (n = 155,206)	0.71 (0.35, 1.45)	0.347	0.52 (0.26, 1.07)	0.078
	≥70 (n = 91,31)	0.84 (0.39, 1.79)	0.645	0.24 (0.09, 0.64)	0.005
Education Level	No formal education (n = 16,34)
	Primary (n = 208, 298)	0.82 (0.29, 2.30)	0.710	0.73 (0.37, 1.45)	0.369
	Secondary (n = 201,126)	0.81 (0.29, 2.27)	0.691	0.50 (0.24, 1.04)	0.064
	College/University (n = 75,42)	0.74 (0.25, 2.20)	0.589	0.42 (0.17, 1.03)	0.058
Occupation	Formal employment (n = 31,40)
	Farming (small and large scale) (n = 78,103)	0.89 (0.39, 2.03)	0.778	0.63 (0.31, 1.28)	0.201
	Self-employed (small and large business) (n = 162,190)	0.92 (0.43, 1.98)	0.828	1.62 (0.84, 3.12)	0.151
	Retired (n = 55,72)	0.58 (0.15, 2.24)	0.432	0.75 (0.36, 1.59)	0.457
	Unemployed (n = 174,92)	0.92 (0.43, 1.98)	0.828	1.43 (0.70, 2.92)	0.322
Family history of diabetes	No (n = 233,205)
	Yes (n = 267,295)	1.03 (0.72, 1.47)	0.867	0.87 (0.62, 1.23)	0.433
Duration from first diagnosis of type 2 diabetes	≥6 years (n = 309,334)
	<6 years (n = 191,166)	0.76 (0.52, 1.09)	0.136	1.22 (0.85, 1.75)	0.286
Health insurance coverage	Uninsured (n = 163,302)
	Insured (n = 337,198)	1.16 (0.80, 1.69)	0.425	0.35 (0.24, 0.50)	<0.001
Presence of co-morbidities	No (n = 141,132)
	Yes (n = 355,368)	0.97 (0.93, 1.02)	0.209	1.00 (0.83, 1.19)	0.965
Economic status (ladder scale 1–10)	Higher SES [6–10] (n = 126,126)
	Lower SES [1–5] (n = 374,374)	0.64 (0.36, 1.13)	0.127	2.58 (1.71, 3.88)	<0.001

Table 4. Predictors of severe disruption of diabetes care in Kenya and Tanzania (multivariate analysis).

		Kenya		Tanzania
Variable	Categories (n=Kenya, Tanzania)	Odds ratio* (95%CI)	p value*	Odds ratio* (95%CI)	p value*
Location	Urban (n=380,300)
	Rural (n=120,200)	0.35 (0.22, 0.58)	<0.001	0.72 (0.45, 1.15)	0.169
Sex	Female (330,336)
	Male (n=170, 164)	1.08 (0.70, 1.67)	0.717	1.08 (0.69,1.68)	0.699
Age group	<40 (n=39,36)
	40–49 (n=80,62)	1.57 (0.71, 3.47)	0.266	0.81 (0.34, 1.94)	0.638
	50–59 (n=135,165)	0.94 (0.44, 1.98)	0.866	0.61 (0.27, 1.34)	0.216
	60–69 (n=155,206)	0.66 (0.31, 1.40)	0.279	0.59 (0.26, 1.34)	0.210
	≥70 (n=91,31)	0.86 (0.38, 1.95)	0.719	0.37 (0.12, 1.17)	0.092
Education Level	No formal education (n=16,34)
	Primary (n=208, 298)	0.68 (0.23, 1.99)	0.487	0.71 (0.34, 1.48)	0.357
	Secondary (n=201,126)	0.66 (0.22, 1.95)	0.454	0.63 (0.27, 1.49)	0.294
	College/University (n=75,42)	0.54 (0.17, 1.72)	0.294	0.48 (0.17, 1.36)	0.165
Occupation	Formal employment (n=31,40)
	Farming (small and large scale) (n=78,103)	1.10 (0.44, 2.79)	0.835	0.48 (0.21, 1.09)	0.080
	Self-employed (small and large business) (n=162,190)	0.76 (0.33, 1.75)	0.515	0.77 (0.36, 1.65)	0.499
	Retired (n=55,72)	0.54 (0.13, 2.20)	0.394	0.81 (0.35, 1.90)	0.630
	Unemployed (n=174,92)	0.88 (0.37, 2.06)	0.763	0.70 (0.30, 1.68)	0.428
Family history of diabetes	No (n=233,205)
	Yes (n=267,295)	1.10 (0.76, 1.58)	0.619	0.97 (0.67, 1.42)	0.881
Duration from first diagnosis of type 2 diabetes	≥6 years (n=309,334)
	<6 years (n=191,166)	0.72 (0.49, 1.06)	0.093	0.93 (0.63, 1.38)	0.731
Health insurance coverage	Uninsured (n=163,302)
	Insured (n=337,198)	1.56 (1.05, 2.34)	0.029	0.51 (0.33, 0.79)	0.003
Presence of co-morbidities	No (n=141,132)
	Yes (n=355,368)	0.97 (0.93, 1.02)	0.301	1.08 (0.89, 1.33)	0.431
Economic status (ladder scale 1–10)	Higher SES [6–10] (n=126,126)
	Lower SES [1–5] (n=374,374)	0.79 (0.43, 1.43)	0.434	1.81 (1.14, 2.88)	0.011

*Adjusted for patient demographic and clinical characteristics (sex, age, education level, occupation, location, family history and duration of diabetes, presence of co-morbidities, economic status and health insurance coverage).

Data availability statement

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.