rss
JAMIA 2008;15:311-320 doi:10.1197/jamia.M2555
  • The Practice of Informatics
  • Review Paper

Prompting Clinicians about Preventive Care Measures: A Systematic Review of Randomized Controlled Trials

  1. Judith W Dexheimera,
  2. Thomas R Talbotb,c,
  3. David L Sandersa,
  4. S Trent Rosenblooma,c,
  5. Dominik Aronskya,d
  1. aDepartment of Biomedical Informatics, Vanderbilt University Medical Center, Nashville, TN
  2. bDepartment of Preventive Medicine, Vanderbilt University Medical Center, Nashville, TN
  3. cDepartment of Medicine, Vanderbilt University Medical Center, Nashville, TN
  4. dDepartment of Emergency Medicine, Vanderbilt University Medical Center, Nashville, TN
  1. Correspondence: Dominik Aronsky, MD, PhD, Department of Biomedical Informatics, Eskind Biomedical Library, Vanderbilt University, 2209 Garland Ave, Nashville, TN 37232-8340 (e-mail: <dominik.aronsky{at}vanderbilt.edu>)
  • Received 12 July 2007
  • Accepted 29 January 2008

Abstract

Preventive care measures remain underutilized despite recommendations to increase their use. The objective of this review was to examine the characteristics, types, and effects of paper- and computer-based interventions for preventive care measures. The study provides an update to a previous systematic review. We included randomized controlled trials that implemented a physician reminder and measured the effects on the frequency of providing preventive care. Of the 1,535 articles identified, 28 met inclusion criteria and were combined with the 33 studies from the previous review. The studies involved 264 preventive care interventions, 4,638 clinicians and 144,605 patients. Implementation strategies included combined paper-based with computer generated reminders in 34 studies (56%), paper-based reminders in 19 studies (31%), and fully computerized reminders in 8 studies (13%). The average increase for the three strategies in delivering preventive care measures ranged between 12% and 14%. Cardiac care and smoking cessation reminders were most effective. Computer-generated prompts were the most commonly implemented reminders. Clinician reminders are a successful approach for increasing the rates of delivering preventive care; however, their effectiveness remains modest. Despite increased implementation of electronic health records, randomized controlled trials evaluating computerized reminder systems are infrequent.

Introduction

The U.S. Preventive Task Force developed guidelines to facilitate the dissemination and implementation of preventive care measures among health care providers.1 2 Opportunities for offering patients preventive care measures exist during most encounters with the health care system,3 such as vaccinations during primary care visits,4 prophylactic aspirin and vaccinations prior to discharge from the hospital,5 or vaccinations during an emergency department visit.6 However, preventive care measures remain underutilized5 7 8 and clinicians struggle with finding time to be compliant with offering the numerous recommended examinations and procedures when a patient's primary visit reason is unrelated to prevention.9 For example, the U.S. Department of Health and Human Services' Healthy People 2010 target for colorectal cancer screening is 50%, but only 35% of eligible people have a screening examination.1 Similarly, the 70% influenza and 65% pneumococcal vaccination rate for patients aged 65 years and older are considerably below the 90% target.10 In a US study examining 38 different preventive care quality indicators patients received only 54.9% of recommended preventive care measures.8

Different implementation approaches to increase preventive care measures have demonstrated various levels of success. Successful approaches include organizational change interventions, financial incentives, or patient and provider reminders.11 12 13 With the increased implementation of clinical information systems, broader adoption and application of information technology for patient care, including preventive care applications, can be expected. In the ambulatory setting computer-based reminders increased the implementation of some preventive care measures, but failed in others.14 15 In an outpatient setting computerized prompts were more effective at increasing influenza vaccination rates when compared to paper-based reminders.16 Balas et al. examined the effect of various intervention techniques for prompting physicians.17 The study included reports from 1966 to 1996 and found that the average rate difference for adherence to recommended preventive care strategies using computer-generated reminders did not differ from non-computerized prompting approaches.

Although a recent US national survey18 suggested that the application of information technology is associated with increased physician reminder use, there is limited information whether the recent focus on implementing clinical information systems has provided the infrastructure to support the development and application of computer-based reminder systems for preventive care. The goal of this systematic literature review was to update the study by Balas et al.,17 which included 16 preventive care measures from the US Preventive Task Force, and to examine whether the amount of computerized reminder systems for preventive care have changed as clinicians increasingly utilize electronic health record systems when providing patient care.

Methods

Literature Search

The study methodology from Balas et al. was adopted to perform a systematic review of the literature regarding 16 preventive medicine reminders to clinicians.17 Eligible studies included randomized controlled trials that targeted clinicians and applied a reminder system for at least one of 16 preventive medicine procedures: fecal occult blood testing; mammography; Papanicolaou smear; influenza, pneumococcal or tetanus vaccination; diabetes mellitus management; cholesterol screening; hemoglobin or blood pressure management; cardiac care; smoking cessation; glaucoma screening; alcohol abuse counseling; prenatal care; and tuberculin testing.

For the period January 1, 1997 to December 31, 2004, we queried the electronic literature databases PUBMED® (MEDLINE®),19 OVID CINAHL®,20 ISI Web of Science™,21 Health and Psychosocial Instruments,20 and the Health Reference Center.22 The search was limited to studies published in English. In each database we searched for the combination of the following three concepts: (1) preventive care measure; (2) reminder system; and (3) randomized clinical trial. In MEDLINE, all search terms were defined as Medical Subject Headings (MeSH®) unless otherwise noted and were searched as they appear below; in the remaining databases, the search terms were defined only as keywords.

  1. Preventive care measure: preventive health services, immunization, vaccination, smoking, smoking cessation, mass screening, mammography, prenatal care, hypertension, blood pressure, diabetes mellitus, alcoholism, substance-related disorders, vaginal smears, hypercholesterolemia, glaucoma, or occult blood.

  2. Reminder system: checklist (text word), encounter forms (text word), tags (text word), triggers (text word), reminder systems, alert (text word), reminder (text word), leaflets (text word), stickers (text word), messages (text word), or tailored messages (text word).

  3. Randomized clinical trial: random' (truncated text word), group' (truncated text word), random allocation, randomized controlled trial (publication type), or clinical trial (publication type).

Review of Identified Studies

The title and abstract of all articles identified using the keyword searches were retrieved and reviewed by two of four independent reviewers (JWD, DLS, SR, DA). Disagreements between two reviewers were resolved by consensus among all four participating reviewers. The bibliographies of identified systematic reviews and meta-analyses were reviewed and additional relevant studies were included. The full text of included articles was obtained and two reviewers (JWD, DA) independently scored each article using the quality assessment instrument that was applied during the previous study.17 Disagreements were resolved by consensus discussion. The assessment instrument included ten criteria evaluating the study characteristics (randomization techniques, testing, withdrawals, effect variables) and assigned a summary score between 0 and 100.23 Five criteria examine the methodology and characteristics of the study design. Following the previous review methodology articles scoring below 50 were excluded from further consideration.23 All included studies were examined for redundancy and duplicate results were removed.

Reminder implementations were classified as “paper-based,” “computer-generated,” or “computerized.” Paper-based reminders included the use of memos, stickers, or a slip of paper within the patient's chart. Computer-generated reminders included application of computerized algorithms to identify eligible patients, but the prompt was printed out and placed in the patient chart to remind the clinician. Computerized reminders included prompts that were entirely electronic, i.e., computerized algorithms identified eligible patients, and prompts were provided upon access to the electronic clinical information system.

Analysis

We combined the articles from the previous review (1966 to 1996)17 with the newly identified articles (1997 to 2004). In studies with more than one preventive care prompt, each intervention was analyzed separately for the effect of the prompt on the given procedure. For example, if a vaccination study compared a paper-based versus a computer-based implementation approach, each approach was counted and examined individually. For each study, the intervention effect was calculated by subtracting the control or baseline data from the largest increase in effect. The unweighted difference in rates of each study was averaged to create the average effect for each implementation strategy, reminder strategy, or intervention. Odds ratios were converted into percentages for data analysis measures. Agreement among reviewers to consider articles based on title and abstract was high (0.96 to 0.99), as determined by Yule's Q.24Graphic

Results

Search Results

The literature search produced 1,535 articles during the time period from 1997 to 2004 (Figure 1). The MEDLINE search contributed 1,308 articles, CINAHL 148, Health and Psychological Instruments three, the Health Reference Center two, and ISI Web of Knowledge 74. After removing 131 duplicate articles, 1,396 were further excluded based on the review of the abstracts. Of the remaining 35 reports, 11 were excluded from further analysis (nine scored less than 50 in the quality assessment, one examined only the system design, and one had no clinician prompt). Reviews of reference lists accounted for an additional 9 articles with 4 meeting inclusion criteria, for a total of 28 included studies. One paper had no numerical results and was not included in the average effect calculations.25 One paper had redundant results and these were removed for analysis.26 We combined the 28 trials with the previous 33 studies for a total of 61 studies.4 5 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 Table 1 shows the detailed characteristics of the included studies.

Figure 1

Flow diagram of included and excluded studies (1997 to 2004).

Table 1

Study Characteristics, Grouped by Implementation Strategy

Number of Studies

Figure 2 displays the number and type of published studies, grouped by intervention's start year. The total number of computerized studies increased in the 2000–2004 period as compared to previous periods. From 2000–2004, nine studies applied paper-based interventions and seven computerized methods, while the number of computer-generated approaches declined to three reports.

Figure 2

Preventive care reminder studies by intervention year. If the intervention's start year was not mentioned in the paper, the publication year − 1 was used to estimate an intervention year.

Interventions

The 61 studies examined a total of 264 preventive care interventions (maximum: 16). Nineteen (30%) studies evaluated three or more preventive measures, three (5%) examined two measures, and the remaining 39 (64%) assessed the impact of one measure. With a total of 110 studied interventions cancer screening (fecal occult blood testing, Papanicolaou smears, and mammograms) was the most frequent type of preventive care measure targeted by clinician reminders, followed by 64 interventions targeting vaccination.

Study Setting

The setting of 33 studies was an academic medical center, while the remaining 28 studies were conducted at non-academic hospitals and clinics. Five studies (9.6%) were performed in an inpatient setting, and the remaining 56 studies were in primary care clinics. In the inpatient setting the delivery of vaccinations were the most frequently studied reminders. The number of facilities in each study ranged from one (39 studies) to 1,655 hospitals or practice groups.

Prompting Clinicians

The methods of prompting clinicians are shown in Table 1. Paper-based combined with computer-generated prompts were the most frequent clinician reminder approach and accounted for 34 studies (52%), followed by 19 paper-based (34%), and 8 computerized studies (13%). The three examined prompting approaches demonstrated a similar average increase in completing preventive care measures (Table 2). Paper-based reminders were applied in 80 interventions and resulted in a 14% average increase of preventive care compliance. Computer-generated reminders were implemented 136 times and had an average increase of 12%. Computerized reminders were employed in 48 interventions and resulted in a 13% average increase.

Table 2

Comparison of Primary Implementation Reminder Strategies

Prompting Methods

Of the 61 studies, 35 examined interventions that prompted only the clinician, 17 combined the clinician prompt with a patient reminder, and 9 studies examined the effects of prompting the clinician in one study group compared to reminding both the clinician and patient in the other study group. To remind patients, 15 mailed reminder letters, and eight studies notified patients via telephone. One study put up fliers and posters for the patients, one study visited patients at their homes to encourage vaccinations, and another study chose to educate patients on the importance of preventive care to encourage return visits. Table 3 summarizes the effectiveness of clinician reminders only, and the combined approach of clinician and patient reminders. The average increase in preventive care procedure compliance was larger when prompting only the clinician (14%) compared to prompting both the clinician and the patient (10%) All but two of the studies prompted the physician before the patient appointment or at the time of order entry.

Table 3

Comparison of Clinician Only versus Combined Clinician–Patient Reminder Strategy

Average Effect

Table 4 displays the effect for measures that were examined by three studies or more. The average effects for measures examined by fewer than three studies ranged from 5% for prenatal care to 14% for alcohol abuse counseling. Prompting clinicians was most effective for smoking cessation (average: 23%), cardiac care (average: 20%), blood pressure screening (average: 16%), followed by vaccinations, diabetes management, and cholesterol (averages: 15%). Mammography reminders had the smallest average effect (10%).

Table 4

Effect of Prompting Clinicians for Preventive Care Procedures in Studies with Three or More Interventions

Discussion

This systematic review summarized findings from 61 randomized controlled clinical trials that investigated the use of reminders to increase preventive care. Overall the prompting of clinicians continued to demonstrate a positive effect on the delivery of the 16 preventive care measures. In recent years, however, the clinician reminder strategies shifted from paper- to computer-based approaches.

Approaches that included a paper-based reminder component (paper-based or computer-generated) remained the most frequent implementation strategies (87%) and had a similar average effect as computerized reminders (14% versus 13%). In studies that included a paper-based component, a reminder sheet is attached on the front of the patient chart or tagged the paper chart in some form, indicating that the paper-record remains an important source of information and documentation instrument in many hospitals and clinics. To implement preventive care measures that require multiple steps during a visit, paper-based solutions may be easier integrated with the clinical workflow as compared to designing an information technology solution that depends on the provider's workstation use. Paper-based implementation strategies are effective when the number of targeted preventive care measures is limited. With an increasing number of recommended preventive care measures, a paper-based process may quickly encounter implementation challenges due to the limited scalability. However, clinical workflow processes that rely on paper charts may continue to favor paper-based implementation strategies.

Computer-generated reminders were most common (52%). In recent years, however, studies on the impact of computer-generated prompts tended to decrease, while computerized reminders increased. The recent increase in applying computerized reminder strategies suggests that clinical information systems are increasingly providing the infrastructure to implement preventive care reminders. Computerized reminder systems require an electronic medical record throughout the practice or hospital; however, only 23.9% of physicians in the US are using an electronic medical record system and 5% of hospitals are using computerized provider order entry systems.84 Implementing preventive care measures using computerized reminders may overcome some of the paper-based implementation challenges. Although clinical information systems may provide an easier to scale and more sustainable infrastructure, they work best when clinicians can complete all steps involved in offering preventive care measures, avoiding the need to switch between paper-based and electronic means. For example, adoption of computerized reminders may be higher if systems apply computerized algorithms for eligibility screening, prompt clinicians at the right time, offer quick ordering processes, and facilitate documentation. Unfortunately the availability of such advanced information system environments remain the exception rather than the rule.

The prompts offered through the reminder systems were heterogeneous. Some preventive care procedures would be easier to perform during the visit, such as vaccinations or blood pressure screening, while others may require a separate appointment, such as mammograms. Cardiac care measures and discussion about smoking cessation were most frequently studied. These preventive measures can be performed during the same visit and are more likely to be completed.

One of the goals of any reminder system is providing the right information at the right time, to the right person, and in the right format. Each study did not report a comprehensive description of the environment and clinical workflow; thus, the effectiveness of a prompt may differ depending on the existing workflow and the effectiveness of a support infrastructure, such as additional personnel or the availability of information technology. Any of the approaches used to implement a reminder system should be examined in combination with the support infrastructure within which the reminder system is implemented.

The review has limitations that may influence the interpretation of the findings. First, it is conceivable that preventive care measures have become an accepted standard of care and further studies examining the effects of reminder strategies are not warranted any more. Many commercial health information systems provide the means of implementing reminders for preventive care measures. By prompting the clinicians, there may be behavioral changes that are not documented as outcomes and may therefore influence the preventive care procedure. Second, the possibility of publication bias exists in the studies we included, there may be less negative studies available. We have not excluded the possibility that publication bias may exist, however, we scored the study design of randomized controlled trials for inclusion in the review. It is possible that only including randomized controlled trials may have excluded some computerized systems which may have used historical controls instead of a concurrent control group. And our choice of using the average difference between baseline and intervention could potentially dilute findings related to specific patient populations. The average differences calculated in each study were not weighted by sample size of patients or clinicians. Lastly, this review only considered studies that involved clinicians in the process of offering patients preventive care measures. It is possible that other implementation strategies exist that do not involve the clinicians. The effectiveness of these approaches may differ in providing preventive care.

The average effect for all of the prompting methods was modest, the act of providing a prompt may modify behaviors that are not measured by the system. The clinicians may have been more aware of the preventive care procedures and may have considered them for more patients. By calculating the average effect by using the historical or concurrent control groups, we sought to make the groups as comparable as possible.

Each encounter with the healthcare system provides an opportunity to offer preventive care measures. However, keeping pace with the many different recommendations and various schedules remains a major challenge for busy clinicians that are expected to focus on a patient's current reason for the visit. Although clinical information systems can keep track of the various recommendations and schedules, they may lead to “prompting fatigue” as an unintended consequence.85 An additional challenge is the fragmentation of health care information,86 which requires providers to repeatedly verify the patient's eligibility for a preventive care measure, a time-consuming task even for one measure. As the healthcare sector applies more information technology, the exchange and sharing of electronic health record information among various providers may lessen that burden in the future.

Conclusion

In summary, this review showed an increase in applying computerized reminder systems for prompting clinicians to offer preventive care measures. As information technology reminder solutions may provide a better scalable and more sustainable model for the increasing burden of following different preventive care guidelines, we saw only a moderate increase in the number of randomized controlled trials looking at computerized reminder systems for preventive care.

Footnotes

  • JWD was supported by training grant from the National Library of Medicine (LM T15 007450-03). STR was supported by a National Library of Medicine grant (LM K22 08576-02).

  • Preliminary results from this work were presented at the 2005 AMIA Annual Symposium (ref: Dexheimer JW, Sanders DL, Rosenbloom ST, Aronsky D. Prompting clinicians: a systematic review of preventive care reminders. AMIA Annu Symp Proc. 2005;:938.)

References

Access policy for JAMIA

All content published in JAMIA is deposited with PubMedCentral by the publisher but with varying embargo times. Authors/funders may pay an Unlocked fee of $2,000 to make the article free on the JAMIA website and PMC immediately on publication. Research funded by government and other recognised agencies is deposited with a 12 month embargo. All other content is deposited with a 36 month embargo.

The Journal of the American Medical Informatics Association is published for the American Medical Informatics Association by BMJ Publishing Group Ltd.