Home > DARE Reviews > Computerized knowledge management in...

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

Database of Abstracts of Reviews of Effects (DARE): Quality-assessed Reviews [Internet]. York (UK): Centre for Reviews and Dissemination (UK); 1995-.

Database of Abstracts of Reviews of Effects (DARE): Quality-assessed Reviews [Internet].

Computerized knowledge management in diabetes care

EA Balas, S Krishna, RA Kretschmer, TR Cheek, DF Lobach, and SA Boren.

Review published: 2004.

CRD summary

This review assessed the effects of computerised information interventions in diabetic care. The authors concluded that prompting follow-up procedures, computerised insulin regimen adjustment using home glucose records, remote feedback and counselling could improve measures of diabetic care. The results were not consistent among the studies or outcomes. The authors' conclusions may be overoptimistic.

Authors' objectives

To assess the effects of computerised information diabetic care interventions on the process and outcomes of diabetic care.


MEDLINE, HealthSTAR, CINAHL, Compendex, Dissertation Abstracts, ABI/INFORM, EBM Reviews: Best Evidence, the Cochrane Database of Systematic Reviews, ERIC, Inspec and PsycINFO were searched. Studies were not restricted by date of publication. The reference lists in reviews and potentially eligible RCTs were handsearched.

Study selection

Study designs of evaluations included in the review

Randomised controlled trials (RCTs) were eligible for inclusion if they had ten or more units of randomisation.

Specific interventions included in the review

Studies that compared computer-assisted diabetic care interventions with no such intervention were eligible for inclusion. Studies of closed-loop insulin delivery systems were excluded.

The review classified interventions as computer-assisted prompting of diabetic care, utilisation of home glucose records in computer-assisted insulin dose adjustment, and computer-assisted diabetes patient education. The mean duration of studies was 7.0 months (+/- 6.2 months).

Participants included in the review

Studies of patients with diabetes mellitus were eligible for inclusion. The included studies were conducted in adults with type I and 2 diabetes and children in both primary care and out-patient settings.

Outcomes assessed in the review

Studies that measured the process or outcomes of diabetic care were eligible for inclusion.

The review assessed process of care using compliance with guidelines for diabetic care and compliance with a list of specified care measures (foot examination, physical examination, glycaemic monitoring, urine protein determination, renal care, cholesterol level, ophthalmologic examination, neurologic care, influenza vaccination and pneumococcal vaccination). The review also assessed individual measures of diabetic care including glycated haemoglobin (HbA1c), patient satisfaction, insulin dose, blood glucose, and cholesterol (the outcomes assessed in the individual studies were reported in tables).

How were decisions on the relevance of primary studies made?

Two reviewers independently selected studies and resolved any disagreements through consensus.

Assessment of study quality

The studies were assessed for 20 criteria based on site, sample, randomisation, process of observation, data quality and statistical analysis (criteria were based on those described by Balas). Studies scoring zero on randomisation, blinding or withdrawals were not included. The authors did not state who performed the validity assessment. The reviewers contacted authors for clarification of ambiguous reporting of methods.

Data extraction

The authors did not state how the data were extracted for the review, or how many reviewers performed the data extraction.

For studies reporting numerous outcomes, a hierarchy of the five most important measures was used to select outcomes for the review, in order: diabetic complications, social functioning, metabolic control, psychological effect, patient satisfaction, patient participation, utilisation of clinical procedures and cost effects of interventions. For each study, the mean difference and 95% confidence interval (CI) between interventions was calculated for HbA1c and blood glucose. For studies reporting a general measure of compliance for specified diabetic care measures, an overall adherence score was calculated by dividing the number of items completed by the number of items recommended.

Methods of synthesis

How were the studies combined?

Where more than three studies provided means and standard deviations and used the same intervention and measured the same outcome, these studies were combined using a random-effects meta-analysis. Pooled mean differences and 95% CIs between interventions were calculated.

How were differences between studies investigated?

The results for meta-analysis were displayed graphically. Differences among the studies were presented in tabular format.

Results of the review

Nine RCTs examined computer-assisted prompting of diabetic care (over 408 clinicians and 4,000 diabetic patients). Twenty-five RCTs examined utilisation of home glucose records in computer-assisted insulin dose adjustment (1,286 adults and 197 children). Ten RCTs examined computer-assisted diabetes patient education (626 patients including 79 children).

Computer-assisted prompting of diabetic care.

Overall compliance with specific guidelines was significantly improved amongst physicians in the intervention group compared with those in the control group. Where a general measure of compliance was provided (4 studies), significant improvement in compliance was found (P<0.05 in 3 of the 4 studies). It was noteworthy that for some outcomes, including HbA1c assessment, foot or ophthalmologic examination, HbAlc, fasting blood sugar, self-measurement of blood glucose and referral to dietary clinic, no differences were detected with regards to compliance.

Utilisation of home glucose records in computer-assisted insulin dose adjustment.

Compared with the control, computer-assisted insulin dose adjustment significantly decreased HbA1c (0.14, 95% CI: -0.11, -0.16), based on 16 studies (2,152 patients), and significantly decreased blood glucose (0.33 mmol/L, 95% CI: -0.28, -0.38), based on 10 studies (594 patients).

Computer-assisted diabetes patient education.

The studies showed statistically significant improvements in HbA1c, pre-lunch blood glucose and serum cholesterol, based on 4 RCTs.

Authors' conclusions

Prompting follow-up procedures, computerised insulin regimen adjustment using home glucose records, remote feedback and counselling could improve measures of diabetic care.

CRD commentary

The review question was clear in terms of the study design, participants, intervention and outcomes. Several relevant sources were searched, but the search terms were not given and the authors did not state whether any language limitations were applied. Two reviewers independently selected studies, which reduces the potential for bias and errors. The methods used to assess validity and extract the data were not described, so it is not known whether any efforts were made to reduce errors and bias. Validity was formally assessed but the results were not reported.

Some relevant information was tabulated. Only studies of similar interventions reporting the same outcome were combined in the meta-analysis. Although statistical heterogeneity was not formally assessed, the results of the meta-analysis were presented in forest plots and these showed no evidence of statistical heterogeneity. Whilst it was clear from the evidence that there had been some success with computerised interventions for diabetic care, the results together with the outcomes were variable. This suggests that the conclusions of the review may be a little optimistic, and that a more detailed examination of the reasons for this variability may be appropriate.

Implications of the review for practice and research

Practice: The authors did not state any implications for practice.

Research: The authors stated that research is required to cost computerised interventions of diabetic care; to assess the effect of computer literacy on access to quality diabetic care by disadvantaged patients; to test comprehensive diabetes clinical management systems on clinical performance and patient outcomes; to examine the effects of replacing conventional visits by telematic contacts; and to assess the potential for internet or hand-held technology in the improvement of diabetic care. They also stated that future studies should assess patient satisfaction.


National Library of Medicine, grant number LM05545; Agency for Health Care Quality, grant number DHHS 5 R01 HS10472-02; Information Society General Directorate of the European Commission project CHS, IST-1999-13352.

Bibliographic details

Balas E A, Krishna S, Kretschmer R A, Cheek T R, Lobach D F, Boren S A. Computerized knowledge management in diabetes care. Medical Care 2004; 42(6): 610-621. [PubMed: 15167329]

Other publications of related interest

This additional published commentary may also be of interest. Anonymous. Computer-assisted care improves outcomes in people with diabetes. Evidence-Based Healthcare Public Health 2004;8:355-6.

Indexing Status

Subject indexing assigned by NLM


Blood Glucose Self-Monitoring; Diabetes Mellitus /blood /drug therapy; Drug Therapy, Computer-Assisted; Hemoglobin A, Glycosylated /analysis; Humans; Insulin /administration & dosage; Patient Compliance; Patient Education as Topic; Physician-Patient Relations; Randomized Controlled Trials as Topic; Reminder Systems; Self Care /methods; Telemetry; Therapy, Computer-Assisted /methods



Database entry date


Record Status

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.

CRD has determined that this article meets the DARE scientific quality criteria for a systematic review.

Copyright © 2014 University of York.

PMID: 15167329