View source for Improving outcomes for high-risk diabetics using information systems

Orzano AJ, Ohman Strickland P, Tallia AF, Hudson S, Balasubramanian B, Nutting PA, Crabtree BF. Improving outcomes for high-risk diabetics using information systems.  J Am Board Fam Med. 2007 May-Jun;20(3):245-51. [PubMed ID: 17548848]

== Question ==

Do simple (electronic and non-electronic) clinical information systems, including systems to identify patients with diabetes (e.g. registries and health risk assessment protocols) and systems to track patients with diabetes (e.g. reminders, checklists, flow sheets, and chart stickers), improve process and outcome measures of quality care?

== Methods ==


===Design===

Cross-sectional study 

===Setting===

50 group and solo New Jersey and Pennsylvania family medicine practices participating in an intervention study aimed at improving adherence to guidelines for multiple conditions.

===Analysis===

Hierarchical logistic regression

===Predictor variables===

Use of clinical information systems, as assessed by clinician self-report, and the presence of comorbid conditions (hypertension and heart disease), as assessed by chart review.  Information systems could be paper- or computer-based. 

===Outcome variables===

American Diabetes Association clinical practice guidelines, with three components: assessment (a process measure), treatment (a process measure), and attainment of targets (an outcome measure).

== Main results ==

<table border="1">
<tr>
<th>ADA Quality Component</th>
<th>Overall adherence</th>
<th>OR (CI) with identification systems</th>
<th>OR (CI) with tracking systems</th>
</tr>

<tr>
<td>Assessment</td>
<td>55%</td>
<td>0.79 (0.55, 1.14)</td>
<td>1.10 (0.80, 1.51)</td>
</tr>

<tr>
<td>Treatment</td>
<td>64%</td>
<td>0.74 (0.87, 1.06)</td>
<td>1.15 (0.87, 1.51)</td>
</tr>

<tr>
<td>Target outcomes (2 of 3)</td>
<td>49%</td>
<td>===1.23 (1.06, 1.44)===</td>
<td>===1.32 (1.11, 1.59)===</td>
</tr>

<tr>
<td>Target outcomes (3 of 3)</td>
<td>11%</td>
<td>1.22 (0.91, 1.63)</td>
<td>1.28 (0.95, 1.73)</td>
</tr>

</table> 
		
OR is per point on a 2-point scale for ''identification systems'' and a 5-pont scale for ''tracking systems''

Among patients with heart disease, ''identification systems'' were associated with meeting targets (OR 2.30, p=.029). 

Among patients with hypertension, i''dentification systems'' were associated with worse adherence to treatment guidelines (OR 0.62, p=.015).  ''Tracking systems'' were associated with better adherence to treatment guidelines (OR 1.27, p=.095), and meeting targets (OR 1.46, p=.027), including the blood pressure target (OR 1.52, p=.0017).

Among patients with neither heart disease nor hypertension, ''identification systems'' improved meeting targets (OR 2.3, p=.029), while ''tracking systems'' improved adherence to assessment guidelines (OR 1.50, p=.030).

== Conclusions ==


Simple clinical information systems to identify and track patients with diabetes appear to improve diabetes care outcomes.  The effect may be especially pronounced in patients with comorbidities such as hypertension and heart disease.

== Commentary ==


Given the increasing burden of chronic disease in an aging population, increasing attention is being paid to improving the quality of chronic disease management.  The widely used Chronic Care Model (1) includes “clinical information systems” as a key component.  This study directly addresses whether clinical information systems, as actually implemented in community settings, delivers on the promise of improving chronic disease management.  

This study provides evidence that simple clinical information systems do improve the quality of care, albeit modestly (with odds ratios in the 1.2 to 1.5 range).  The American Diabetes Association quality measures are standards with broad support.  Although process measures did not appear to benefit from clinical information systems in the same manner as outcome measures, the authors rightly point out that outcome measures are more likely to be clinically relevant.

The most interesting aspect of this study lies in the authors’ commentary about the relative value of non-computerized clinical information systems vs. computerized clinical information systems.  In the discussion, the authors refer to a second study they performed (2), which suggested that “practices with identification systems (if used in conjunction with an EHR) seem to do better on patient care despite the EHR, rather than because of any advantage provided by the EHR.”  This does not imply that EHRs are without value, but it is consistent with other reports showing that EHRs are not a panacea (3), and information systems do not need to be electronic to have value.

===Stephen E. Ross, MD===

University of Colorado at Denver and Health Sciences Center

Aurora, Colorado, USA

== References ==


# Bodenheimer T, Wagner EH, Grumbach K.  Improving primary care for patients with chronic illness: the chronic care model, Part 2.  JAMA 2002; 288:1909-14.
# Crosson JC, Ohman-Strickland PA, Hahn KA, DiCicco-Bloom B, Shaw E, Orzano AJ, Crabtree BF.  Electronic medical records and diabetes quality of care: results from a sample of family medicine practices.  Ann Fam Med 2007; 5:209-15
# Linder JA, Ma J, Bates DW, Middleton B, Stafford RS.  Electronic heath record use and the quality of ambulatory care in the United States.  Arch Intern Med 2007; 167:1400-1405.

[[Category: BMI-512-F-07]]