Diabetes Research and Clinical Practice 74 (2006) 90–94

www.elsevier.com/locate/diabres

The influence of glucose self-monitoring on glycaemic

control in patients with diabetes mellitus in Sudan
M. Abdelgadir a,b,*, M. Elbagir b, M. Eltom b, C. Berne a
a
Department of Medical Sciences, Uppsala University Hospital, Sweden
b
Endocrinology and Diabetes Research Centre, Omdurman Teaching Hospital, Omdurman, Sudan
Received 24 March 2005; received in revised form 12 February 2006; accepted 7 March 2006
Available online 18 April 2006

Abstract
Objective: To investigate the influence of self-monitoring of glucose on the glycaemic control in Sudanese diabetic subjects.
Subjects and methods: A group of 193 consecutive type 2 and type 1 diabetic subjects (95 men, 98 women) were studied. In 104
subjects with type 2 diabetes fasting blood glucose was measured using a glucose meter and blood was obtained for serum glucose
measurement in the laboratory. In the remaining 89 diabetic subjects random blood glucose was measured using the same glucose
meter and a whole blood sample was drawn for laboratory assessment of HbA1c. Data on self-monitoring and other clinical and
personal characteristics were recorded.
Results: More than 75% of either type 1 and type 2 diabetic patients never self-monitored blood or urine glucose. In type 2
diabetic subjects self-monitoring of blood or urine glucose was not related to glycaemic control. In type 1 diabetic subjects,
however, self- monitoring of blood glucose was significantly associated with better glycaemic control, as assessed by HbA1c (P
= 0.02) and blood glucose at clinic visits (P 0.0001), and similar associations were found for urine glucose self-monitoring (P =
0.04 and 0.02) respectively.≤ Neither glycaemic control nor glucose self-monitoring was associated with education level.
Conclusions: Self-monitoring of blood glucose was not found to be associated to better glycaemic control in Sudanese subjects
with type 2 diabetes. In contrast, self-monitoring of both blood and urine glucose was significantly associated with glycaemic
control in subjects with type 1 diabetes. Self-monitoring of urine glucose could be useful where measurement of blood glucose is
not available or affordable.
# 2006 Elsevier Ireland Ltd. All rights reserved.
Keywords: Diabetes mellitus; Self-monitoring; Sudan

1. Introduction valence was estimated to be 3.4%, whereas the

prevalence was higher in certain communities in the
The prevalence of diabetes mellitus (DM) in the Northern State, where it reached 10.4% [1,2]. Approxi-
Sudan, as in many other low-income countries, is mately, 75% of the diabetic patients have type 2
increasing to epidemic proportions. Diabetes is com- diabetes, and 40% of them were obese and had a
mon among the adult population and diabetes pre- family history of diabetes [3]. DM in this population is
often characterized by poor glycaemic control, with a
high prevalence of acute and chronic complications and
* Corresponding author at: Clinical Research Department 2, Lab 16, a low quality of life [4–7]. Due to limited resources,
Uppsala University Hospital, SE-751 85 Uppsala.
Tel.: +46 18 6113298; fax: +46 18 553601. most of the patients (51%) had reduced or abandoned
E-mail address: moawia.abdelgadir@medsci.uu.se insulin therapy due to non-availability or non-
(M. Abdelgadir). affordability of this drug. Most patients did not
receive a satisfactory
0168-8227/$ – see front matter # 2006 Elsevier Ireland Ltd. All rights
reserved. doi:10.1016/j.diabres.2006.03.003
 M. Abdelgadir et al. / Diabetes Research and Clinical Practice 74 (2006) 90–94 9
diabetes care and education, leading to lower rate of ence range of 3.5–5.3%. All samples were stored and trans-
clinic attendance (55%), and dietary non-compliance 20 8C until analysed at Uppsala University
ported frozen at —
(78.5%) [8]. Hospital, Sweden. HbA1c was determined by high perfor-
Self-monitoring of blood glucose has become of mance liquid chromatography and glucose using a glucose
prime importance in management of both type 1 and oxidase technique in the hospital’s routine laboratory
type 2 diabetic patients [9–11]. A variety of glucose service.
meters have been used with varying accuracy and
precision [12–14]. The cost of self-monitoring of 3. Statistical analysis
blood glucose (SMBG) is a matter of concern, also
among patients residing in high-income countries [15], Statistical analysis was performed using the
result- ing in decreased frequency of monitoring [16]. program Statview for Windows 5.0.1 (SAS Institute
Diabetic patients generally have 2–3 times higher Inc., Cary NC). Results are given
T in mean S.D. unless
health care costs than non-diabetic subjects [17–19]. otherwise specified.
For both type 1 and type 2 diabetes, the clinical All values obtained by the glucose meter device
course and prognosis of the disease are closely linked were multiplied by 1.12 to adjust for the difference
to the glycaemic control and duration of diabetes [20]. between the whole blood and serum values [12].
Thus, adherence to the therapeutic regimens as Wilcoxon Signed Rank was used to compare between
outlined in recent recommendations on intensive variables.
glycaemic control [9,11,21] could potentially result in a
substantial decrease in diabetes complications. 4. Results
The aim of this study was to examine the influence
of self-monitoring of glucose on the glycaemic control The demographic and clinical characteristics of the
in Sudanese subjects with diabetes mellitus. diabetic subjects are shown in Table 1.
There was a significant correlation between blood
glucose values measured with the glucose meter and
2. Subjects and methods values determined in the laboratory (r = 0.77; P =
0.0001).
In this cross-sectional study, a group of 193 consecutive
type 2 (n = 143) and type 1 (n = 50) diabetic subjects (95 The frequency of self-monitoring of blood glucose
men, 98 women) were recruited from the outpatient diabetes (SMBG) and self-monitoring of urine glucose (SMUG)
clinic, Omdurman Teaching Hospital. Inclusion criteria were related to the mean blood glucose are shown in Table
age 20 years and duration of diabetes 1 year. Informed 2. The glycaemic control in the type 2 diabetic subjects
≤ consent was obtained from all ≤ subjects and the study was did not differ significantly by frequency of SMBG or
approved by the health authorities. A questionnaire including SMUG when assessed as HbA1c. Patients who
personal details, clinical characteristics and data on practiced
patients glucose

Table 1
Demographic and clinical characteristics of the diabetic subjects
self-monitoring was completed for all subjects. Blood pres-
sure, weight in kilograms and height in meters in light Characteristics
clothing without shoes were measured for calculation of
body mass index (BMI). Number of subjects 193
The patient’s education level was classified into four Males/females 95/98
categories. The first category was no education, the second Duration of diabetes (years) 10.1 T 7.9
primary education, the third secondary education, and the Age (years) 50.0 T 13.4
fourth comprised those with university education or higher. BMI (Kg/m2) 22.9 T 4.9
To assess acute glycaemic control, fasting blood glucose Systolic BP (mmHg) 122.7 T 15.3
was measured using portable glucose meters Accutrend1 Diastolic BP (mmHg) 81.6 T 9.2
Sensor (F. Hoffmann-La Roche, Basle, Switzerland) from
104 type 2 diabetic subjects (45 males, 59 females). A Mean T S.D.
venous blood sample was drawn immediately thereafter and
centri- Table 2
Frequency distribution of SMBG and SMUG related to blood glucose
fuged and serum was separated within 2 h for laboratory
measurement of glucose. Self-monitoring SMBG SMUG
Random blood glucose was measured in the remaining 89 technique Blood glucose (mmol/l)
diabetic subjects (50 males, 39 females), 39 of whom had Once a day 6.2 T 1.8 (n = 4) 7.4 T 2.9 (n = 7)
type 2 diabetes and 50 had type 1 diabetes using the same Once a week 9.4 T 3.5 (n = 48) 10.5 T 3.5 (n = 33)
device, and venous whole blood was drawn in EDTA- None 13.1 T 4.5 (n = 141) 13.4 T 4.5 (n = 153)
containing tubes for determination of haemoglobin A1c
(HbA1c) with a refer- Mean T SD.
SMUG daily had HbA1c 7.2 T 1.2%, once weekly 5. Discussion
7.6 T 1.9% or never 9.3 T 2.3%. Those who practised
SMBG weekly had HbA1c 9.1 T 2.2% and never In this report from an urban population in Sudan we
7.1 T 1.9%. A similar finding was observed in type 2 have shown that the frequency of self-monitoring of
diabetic subjects assessed by fasting serum glucose. glucose (blood or urine) was positively associated to
Also the patients’ level of education appeared to good glycaemic control in type 1 diabetes but not in
be unrelated to glycaemic control, measured by fasting type 2 diabetes patients. Education level of the patients
serum glucose, in subjects with type 2 diabetes. Those was neither associated to frequency of self-monitoring
who had no education had 8.1 4.3 mmol/l as nor to level of glycaemic control.
compared to 9.1 5.0 mmol/l among T those with
The frequency of glucose self-monitoring in type 2
T
primary education, 7.8 3.6 mmol/l among those with diabetic patients in this study was considerably lower
secondary education and T 8.9 3.8 mmol/l when having than recommended [12], albeit the effectiveness of
T
university education. Further, education was not SMBG on glycaemic control for type 2 diabetic
related to frequency of SMUG. The same observation subjects is debatable [9,22–25]. However, a recent
was made in type 2 diabetes subjects assessed by consensus suggests that, despite doubts of the evidence
HbA1c. Education was not found to correlate with of the effects of glucose self-monitoring on glycaemic
either glycaemic control or patterns of both SMBG and control in type 2 diabetes, in special circumstances
SMUG. such as shifting to a new oral hypoglycaemic agent,
Eighty four percent of subjects with type 1 diabetes during periods of illness, to document post-prandial
(n = 42) never performed SMUG, 14% (n = 7) did so hypergly- caemia and during administration of
once a week and only 2% (n = 1) once daily. Eighty systemic gluco- corticosteroids, routine self-monitoring
percent did not perform SMUG (n = 40), while 10% is warranted [26].
did soonce per week (n = 5) and 10%(n = 5) once In type 1 diabetes intensive insulin therapy and
daily. Type 1 diabetic subjects, who never monitored intensive management are generally considered impor-
blood glucose, had significantly higher random blood tant, including SMBG [27]. In this study, the frequency
glucose than those who did (17.2 4.5 mmol/l of SMBG was significantly associated to better
versus 7.2 1.8Tmmol/l, T glycaemic control in type 1 diabetes, supporting prior
P ≤ 0.0001), and higher HbA1c (9.4 2.1% versus reports. This emphasises potential serious long-term
5.6 T1.5%, P = 0.02). Type 1 diabetic patients who consequences, when patients in a low-income country
never monitored urine glucose had also significantly setting, as a consequence of infrequent self-monitoring,
increased random glucose levels compared to those get a poor glycaemic control with ensuing early
who did at least once weekly (17.2 4.8 mmol/l diabetic complications [5,8,28].
versus T Urine glucose monitoring is less frequently recom-
8.3 T2.3 mmol/l, P = 0.02), and higher HbA1c (9.4 T mended as a tool for glycaemic control [27]. Still,
2.2% versus 6.7 T1.3%, P = 0.04). SMUG could be useful and cost-effective where no
Similarly, there was no significant association other alternatives are available [29,30]. Despite receiv-
between education and HbA1c in type 1 diabetic ing advice on how to divide each urine strip long-
subjects; 9.2 T 3.1%, 8.6 T 2.0%, 8.9 T2.4%, and itudinally to reduce the cost and to increase frequency
9.7 T2.0% for no education, primary, secondary and of urine self-monitoring, a high percentage of patients
university education, respectively. Education in the in our study never performed SMUG, which resulted in
type
1 diabetic patients was not associated with the
frequency of SMUG (Table 3).

Table 3
HbA1c (%) for patients with different levels of education
Education SMBG
Daily Weekly None
None (n = 8) 5.6 T 2.4 (n = 2) 10.4 T 2.4 (n = 6) a

Primary (n = 19) 5.3 T 0.0 (n = 1) 7.1 T 1.3 (n = 2) 9.0 T 1.8 (n = 16)

Secondary (n = 20) a
6.3 T 1.3 (n = 3) 9.4 T 2.3 (n = 17)
University (n = 3) a a
8.9 T 2.4 (n = 3)
Results indicated for different degrees of blood glucose self-monitoring
a
No observations.
significantly increased HbA1c and glucose levels glucose tolerance in the Danagla community in northern Sudan,
compared to those performed testing. Diabet. Med. 15 (2) (1998) 164–169.
The majority of both type 1 and type 2 diabetic [3] E.M.A. Elmahadi, A.R. Salih, E. Mukhtar, Patterns of diabetes
patients in this Sudanese population did not practice mellitus in the Sudan, Trop. Geogr. Med. 4 (1989) 353–355.
[4] A. Elamin, A. Eltahir, B. Ismail, T. Tuvemo, Clinical pattern of
any kind of SMUG. The high cost of the devices as childhood Type 1 (insulin dependent) diabetes mellitus in the
well as the high running cost of self-monitoring may be Sudan, Diabetelogia 35 (1992) 645–648.
the primary factor behind this problem. The lack of [5] M.N. Elbagir, M.A. Eltom, E.O. Mahadi, C. Berne, Pattern of
association between education level of the patients long-term complications in Sudanese insulin-treated diabetic
and self-monitoring frequency and glycaemic control patients, Diabetes Res. Clin. Pract. 30 (1) (1995) 59–67.
[6] M.N. Elbagir, N.O. Etayeb, M.A. Eltom, E.O. Mahadi, K.
indicates that most likely social group as well as Wikblad, C. Berne, Health-related quality of life in insulin-
economic resources do not explain the variation within treated diabetic patients in the Sudan, Diabetes Res. Clin. Pract.
the group. More important may be diabetes education 46 (1) (1999) 65–73.
and raised patient awareness, both of which are likely [7] M. Abdelgadir, M. Elbagir, A. Eltom, M. Eltom, C. Berne,
to have an impact both upon rates of self-monitoring Factors affecting perinatal morbidity and mortality in pregnan-
cies complicated by diabetes mellitus in Sudan, Diabetes Res.
and glycaemic control [6,9]. Clin. Pract. 60 (1) (2003) 41–47.
For the Sudanese type 1 diabetic patients, manage- [8] M.N. Elbagir, M.A. Eltom, H. Rosling, C. Berne, Glycaemic
ment of diabetes represents a real burden, and insulin control of insulin-dependent diabetes mellitus in Sudan: influ-
cost is a serious problem [31,32], in addition to cost of ence of insulin shortage, Diabetes Res. Clin. Pract. 30 (1)
(1995) 43–52.
self-monitoring. For type 2 diabetes patients, on the [9] J.N. Sarol, N.A. Nicodemus, K.M. Tan, M.B. Grava, Self-
other hand, self-monitoring appeared not to be cost- monitoring of blood glucose as part of a multi-component
effective. Oral hypoglycaemic agents are locally therapy among non-insulin requiring type 2 diabetes patients:
produced, affordable and available, and that may partly a meta-analysis, Curr. Med. Res. Opin. 21 (2) (2005) 173–
explain the relatively better glycaemic control as 184.
[10] S.B. Soumerai, C. Mah, F. Zhang, A. Adams, M. Barton, V.
compared to type 1 diabetic patients. Fajtova, et al., Effects of health maintenance organization
In conclusion, we have shown that self-monitoring cover- age of self-monitoring devices on diabetes self-care and
of both blood and urine glucose is not widely practiced glyce- mic control, Arch. Intern. Med. 164 (6) (2004) 645–
among Sudanese diabetic patients, probably due to 652.
[11] American Diabetes Association, Standards for medical care for
high cost and insufficient diabetes education. Self- patients with diabetes mellitus, Diabetes Care 28 (Suppl. 1)
monitor- ing in type 2 diabetic patients had little or no (2005) S4–S36.
effect on glycaemic control, while, in contrast, self- [12] J.Y. Poirier, N. Le Prieur, L. Campion, I. Guilhem, H. Allannic,
monitoring was strongly linked to the glycaemic D. Maugendre, Clinical and statistical evaluation of self-mon-
control in type 1 diabetic patients. Diabetes education itoring blood glucose meters, Diabetes Care 21 (1998) 1919–
and improving of health services at the primary level 1924.
[13] R. Lehmann, S. Kayrooz, H. Greuter, G.A. Spinas, Clinical and
are important for the reduction of complications and technical evaluation of a new self-monitoring blood glucose
burden of disease for the increasing group of Sudanese meter: assessment of analytical and user error, Diabetes Res.
diabetes patients. Clin. Pract. 53 (2001) 121–128.
[14] R.N. Johnson, J.R. Baker, Analytical error of home glucose
monitors: a comparison of 18 systems, Ann. Clin. Biochem. 36
Acknowledgements (1999) 72–79.
[15] A.J. Karter, A. Ferrara, J.A. Darbinian, L.M. Ackerson, J.V.
We thank the staff of diabetes clinic, Omdurman Selby, Self-monitoring of blood glucose: language and financial
barriers in a managed care population with diabetes, Diabetes
Teaching Hospital, Sudan and the diabetic subjects for Care 23 (4) (2000) 477–483.
co-operation. [16] M.G. Kibriya, L. Ali, N.G. Banik, A.K. Khan, Home
This work was supported by grants from In-Develop monitoring of blood glucose (HMBG) in type 2 diabetes
Uppsala and the Swedish Diabetes Association. mellitus in a developing country, Diabetes Res. Clin. Prac. 46
(3) (1999) 253–257.
[17] R.J. Rubin, M.W. Altman, D.N. Mendelson, Health care expen-
Reference ditures for people with diabetes milletus, J. Clin. Endocrinol.
Metab. 78 (1994) 809A–809F.
[1] M.N. Elbagir, M.A. Eltom, E.M. Elmahadi, I.M. Kadam, C. [18] P. Clarke, A. Gray, R. Legood, A. Briggs, R. Holman, The
Berne, A population-based study of the prevalence of diabetes impact of diabetes-related complications on healthcare costs:
and impaired glucose tolerance in adults in northern Sudan, results from the United Kingdom Prospective Diabetes Study
Diabetes Care 19 (10) (1996) 1126–1128. (UKPDS Study No. 65), Diabet. Med. 20 (6) (2003) 442–450.
[2] M.N. Elbagir, M.A. Eltom, E.M. Elmahadi, I.M. Kadam, C. [19] F. Henriksson, C.D. Agardh, C. Berne, J. Bolinder, F. Lonnqvist,
Berne, A high prevalence of diabetes mellitus and impaired P. Stenstrom, et al., Direct medical costs for patients with type 2
diabetes in Sweden, J. Intern. Med. 248 (5) (2000) 387–396.
[20] I.M. Stratton, A.I. Adler, H.A. Neil, D.R. Matthews, S.E. Man-
[26] D. Owens, A.H. Barnett, J. Pickup, D. Kerr, P. Bushby, D.
ley, C.A. Cull, et al., Association of glycaemia with macrovas-
Hicks, et al., Blood glucose self-monitoring in type 1 and type
cular and microvascular complications of type 2 diabetes
2 diabetes: reaching a multidisciplinary consensus, Diabet
(UKPDS 35): prospective observational study, Br. Med. J.
Primary Care 6 (2004) 8–16.
12;321 (7258) (2000) 405–412. [27] American Diabetes Association, Tests of glycaemia in diabetes,
[21] The Diabetes Control and Complication Trial Research Group,
Diabetes Care 27 (Suppl 1) (2004) S91–S93.
The effect of intensive treatment of diabetes on the [28] J.H. Devries, F.J. Snoek, R.J. Heine, Persistent poor glycaemic
development and progression of long-term complications in
control in adult Type 1 diabetes. A closer look at the problem,
insulin-dependent diabetes mellitus, N. Engl. J. Med. 329 (14)
Diabet. Med. 21 (12) (2004) 1263–1268.
(1993) 977–986. [29] A.P. Rotchford, K.M. Rotchford, T. Machattie, G.V. Gill, Asses-
[22] R. Weitgasser, B. Gappmayer, M. Pichler, Newer portable
sing diabetic control – reliability of methods available in
glucose meters – analytical improvement compared with pre-
resources poor settings, Diabet. Med. 19 (3) (2002) 195–200.
vious generation device, Clin. Chem. 45 (10) (1999) 1821– [30] E.G. Starostina, M. Antsiferov, G.R. Galstyan, C. Trautner, V.
1825.
Jorgens, U. Bott, et al., Effectiveness and cost-benefit analysis
[23] L.M. Welschen, E. Bloemendal, G. Nijpels, J.M. Dekker, R.J.
of intensive treatment and teaching programmes for type 1
Heine, W.A. Stalman, et al., Self-monitoring of blood glucose
(insulin- dependent) diabetes mellitus in Moscow–blood
in patients with type 2 diabetes who are not using insulin: a
glucose versus urine glucose self-monitoring, Diabetologia 37
systematic review, Diabetes Care 28 (6) (2005) 1510–1517.
(2) (1994) 170– 176.
[24] S. Coster, M.C. Gulliford, P.T. Seed, J.K. Powrie, R. Swami- [31] M. Elmahadi, A. Kaballo, E. Mukhtar, Features of NIDDM in
nathan, Self-monitoring in Type 2 diabetes mellitus: a meta- the Sudan, Diabets Res. Clin. Prac. 11 (1991) 59–63.
analysis, Diabet. Med. 17 (11) (2000) 755–761. [32] H. Elrayah, M. Eltom, A. Bedri, A. Belal, H. Rosling, C.G.
[25] M. Franciosi, F. Pellegrini, G. De Berardis, M. Belfiglio, D.
Ostenson, Economic burden on families of childhood type 1
Cavaliere, B. Di Nardo, et al., The impact of blood glucose self- diabetes in urban Sudan, Diabetes Res. Clin. Pract. 70 (2)
monitoring on metabolic control and quality of life in type 2 (2005) 159–165.
diabetic patients: an urgent need for better educational
strategies, Diabetes Care 24 (11) (2001) 1870–1877.