SUPPORT Summaries
Evidence of the effects of health system interventions
for low- and middle-income countries
Evidence of the effects of health system interventions
for low- and middle-income countries
September, 2009 - SUPPORT Summary of a systematic review | print this article |
Drug therapy often does not provide the intended benefits to patients. Doctors may prescribe the wrong drug or dosage and patients do not always understand the instructions they are given. One potential approach to reducing medication errors is to use a computerised decision support system. These systems can use existing electronic data from, for example, an electronic medical record, to give reminders (delivered at the time of, or before, decision-making) or feedback (generally aggregate information frommultiple patients with the intention of altering future decisions) to healthcare providers or patients. Key messages Computerised reminders to healthcare providers probably improve the managementof patient medication in outpatient settings. Computerised reminders to healthcare providers may not lead to any difference inthe management of patient medication in inpatient settings. It is not known whether reminders directed at patients lead to an increase in patients taking medication as prescribed. The applicability of computerised reminders and feedback depends on the availability of computerised patient record systems for clinicians and on patients’ access to telephones or mobile phones.
Drug therapy often does not provide the intended benefits to patients. Doctors may prescribe the wrong drug or dosage and patients do not always understand the instructions they are given. One potential approach to reducing medication errors is to use a computerised decision support system. These systems can use existing electronic data from, for example, an electronic medical record, to give reminders (delivered at the time of, or before, decision-making) or feedback (generally aggregate information from multiple patients with the intention of altering future decisions) to healthcare providers or patients.
A summary of studies addressing a clearly formulated question that uses systematic and explicit methods to identify, select, and critically appraise the relevant research, and to collect and analyse data from the included studies.
SUPPORT – was an international project to support the use of policy relevant reviews and trials to inform decisions about maternal and child health in low- and middle-income countries, funded by the European Commission (FP6) and the Canadian Institutes of Health Research.
Medication errors such as prescribing the wrong drug or dosage, or patients failing to take a drug as prescribed, can lead to failure of a treatment plan, increased morbidity or death. Computerised decision support systems that give reminders or feedback to healthcare providers or patients are one approach to reducing these errors. Such systems are increasingly viable with growing numbers of providers having access to computerbased patient records, and patients having access to mobile telephones.
The reminder and feedback systems described in the review were used in the care of patients with a wide range of diseases and conditions.
Note that this review excluded systems that calculate drug doses for clinicians as this topic is already covered by another Cochrane review (see related literature on last page).
| / | What the review authors searched for | What the review authors found |
|---|---|---|
| Interventions | Randomised controlled trials (RCTs) of computer generated medication reminders or feedback directed at healthcare providers or patients. |
26 RCTs. Interventions included reminders on usage of medication, electronic prescribing guidelines, point-of-care evidencebased messages, feedback on brand versus generic prescribing, specific reminders at patient visits and telephone repeatprescribing reminders. |
| Participants | Healthcare providers and patients. | 23 studies described interventions aimed at providers, usually general medicine or general practice doctors, although nurses were mentioned in five studies. Three studies described reminder interventions aimed at patients. |
| Settings | Any | Settings were diverse although most were in general medicine (12) or general practice (8). Inpatient and outpatient clinics were included. Studies came from the USA (21), UK (4) and Norway (1). |
| Outcomes | Primary outcome: medication use Secondary outcomes: non-clinical outcomes such as rate of generic prescribing or costs. | 19 trials evaluated the effect of a computerised decision support system on provider adherence to medication guidance while 3 considered patient adherence. 3 studies assessed cost as their main outcome. |
| Date of most recent search: December 2001 | ||
| Limitations: This is a good quality systematic review with only minor limitations. The review is, however, old given the nature of information technology and this should be borne in mind when interpreting this summary. | ||
High: This research provides a very good indication of the likely effect. The likelihood that the effect will be substantially different† is low.
Moderate: This research provides a good indication of the likely effect. The likelihood that effect will be substantially different† is moderate.
Low: This research provides some indication of the likely effect. Howevver, the likelihood that it will be substantially different† is high.
Very low: This research does not provide a reliable indication of the likely effect. The likelihood that the effect will be substantially different† is very high.
* This is sometimes referred to as 'quality of evidence' or 'confidence in the estimate'.
† Substantially different = a large enough difference that it might affect a decision.
See last page for more information.
The review included 26 studies reporting 29 comparisons. The wide variety of studies in diverse settings (though most were in general medicine or general practice) meant that it was not possible to combine data from different studies to provide an overall estimate of effect. The comparisons were grouped into five categories and four are presented here. The fifth category, combined reminders and feedback to healthcare providers, comprised a single study from 1986 that suggested that the addition of feedback to reminders provided no benefit. None of the studies was conducted in a low or middle-income country.
Reminders to healthcare providers in outpatient settings Twelve of the 26 studies used a variety of methods to remind healthcare providers about some aspect of a patient’s medication management in outpatient settings. The types of reminders included electronic guidelines, educational messages about individual patients and systems to monitor prescriptions. Poor user interface and requirements that healthcare providers enter redundant data were cited as reasons for poor system utilization.
Computerised reminders to healthcare providers in outpatient settings probably improve the management of patient medication.
No information on patient outcomes (e.g. control of diabetes, or depression symptoms) was presented.
Twelve of the 26 studies used a variety of methods to remind healthcare providers about some aspect of a patient’s medication management in outpatient settings. The types of reminders included electronic guidelines, educational messages about individual patients and systems to monitor prescriptions. Poor user interface and requirements that healthcare providers enter redundant data were cited as reasons for poor system utilization.
Beta-lactam antibiotics compared to Bata-lactam + Aminoglycoside for the treatment of suspected late onset sepsis in new born infants |
|||||
| Patient or population: patients in treatment for suspected late onset neonatal sepsis Settings: High income countries Intervention: Beta-lactam monotherapy (Timentin) Comparison: Bata-lactam + Aminoglycoside combination therapy (Flucloxacilin + Gentamicin) |
|||||
| Outcomes | Impact | Relative effect (95% CI) |
No of Participants (studies) |
Quality of the evidence (GRADE) |
Comments |
| Mortality prior to discharge |
182 per 1000 |
31 per 1000 (2 to 588) |
RR 0.17 |
24 (1) |
1,2 |
| Treatment failure | 182 per 1000 | 31 per 1000 (2 to 588) | RR 0.17 (0.01 to 3.23) |
24 (1) |
1,2 |
| CI: Confidence interval; RR: Risk ratio GRADE: GRADE Working Group grades of evidence (see explanations) | |||||
| 1 No information on method of randomization is provided. Allocation concealment is unclear. Blinding is not documented for the intervention or outcome. 2 Estimate from only one study with few events. Very wide confidence interval consistent with either important harms or important benefits from the therapy |
|||||
Twelve of the 26 studies used a variety of methods to remind healthcare providers about some aspect of a patient’s medication management in outpatient settings. The types of reminders included electronic guidelines, educational messages about individual patients and systems to monitor prescriptions. Poor user interface and requirements that healthcare providers enter redundant data were cited as reasons for poor system utilization.
Provider reminders in outpatient settings |
||||
| Patient or population: Healthcare providers Settings: Outpatient clinics in primary or secondary care Intervention: Reminders Comparison: No reminders |
||||
| Outcomes | Absolute change1 (range) |
Relative effect (95% CI) |
No of studies | Quality of the evidence (GRADE) |
| Appropriate prescribing |
10% increase to 34% increase |
1.0 to 42.0) |
12 |
|
| Patient outcomes | No data presented | No data presented | 0 | - |
| CI: Confidence interval; RR: Risk ratio GRADE: GRADE Working Group grades of evidence (see explanations) | ||||
| 1 Only 8 of the 12 studies provided these data. | ||||
| Findings | Interpretation* |
|---|---|
| APPLICABILITY | |
|
Applicability of the findings will depend on the availability of computerised patient record systems for clinicians and on patients’ access to telephones or mobile phones.
Reminders and feedback assume that poor medication management is due to a lack of knowledge or forgetfulness. Where the problem is poor physical or financial access to medications or health facilities, reminders or feedback are unlikely to be effective.
In some situations a reminder may require a long-term relationship between clinician and patient to be effective. Such relationships may be uncommon in some low and middle-income country settings.
|
| EQUITY | |
|
|
| ECONOMIC CONSIDERATIONS | |
|
|
| MONITORING & EVALUATION | |
|
|
| *Judgements made by the authors of this summary, not necessarily those of the review authors, based on the findings of the review and consultation with researchers and policymakers in low and middle-income countries. For additional details about how these judgements were made see: http://www.support-collaboration.org/summaries/methods.htm | |
Related literature
Durieux P, Trinquart L, Colombet I, Niès J, Walton RT, Rajeswaran A, Rège Walther M, Harvey E, Burnand B. Computerized advice on drug dosage to improve prescribing practice. Cochrane Database of Systematic Reviews 2008, Issue 3. Art. No.: CD002894. DOI: 10.1002/14651858.CD002894.pub2. Go to Cochrane Library | Go to epistemonikos
de Jongh T, Gurol-Urganci I, Vodopivec-Jamsek V, Car J, Atun R. Mobile phone messaging telemedicine for facilitating self management of long-term illnesses (Protocol). Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No.: CD007459. DOI: 10.1002/14651858.CD007459. Go to Cochrane Library
This summary was prepared by
Shaun Treweek, Norwegian Knowledge Centre for the Health Services, Oslo, Norway.
Conflict of interest
None declared. For details, see: www.support-collaboration.org/summaries/coi.htm
Acknowledgements
This summary has been peer reviewed by: Shahed Hossain, Bangladesh; Yannis Pappas, UK; Tomas Pantoja, Chile; Chris Seebregts, South Africa.
This summary should be cited as
Treweek S. Are computerised reminders and feedback to support medication management effective? A SUPPORT Summary of a systematic review. September 2009.www.supportcollaboration. org/summaries.htm
