A patient decision aid to support shared decision-making on anti-thrombotic treatment of patients with atrial fibrillation: randomised controlled trial

The management of non-valvular atrial fibrillation (NVAF) in Australian general practice: bridging the evidence-practice gap. A national, representative postal survey

Background

General practitioners (GPs) are ideally placed to bridge the widely noted evidence-practice gap between current management of NVAF and the need to increase anticoagulant use to reduce the risk of fatal and disabling stroke in NVAF. We aimed to identify gaps in current care, and asked GPs to identify potentially useful strategies to overcome barriers to best practice.

Methods

We obtained contact details for a random sample of 1000 GPs from a national commercial data-base. Randomly selected GPs were mailed a questionnaire after an advance letter. Standardised reminders were administered to enhance response rates. As part of a larger survey assessing GP management of NVAF, we included questions to explore GPs' risk assessment, estimates of stroke risk and GPs' perceptions of the risks and benefits of anticoagulation with warfarin. In addition, we explored GPs' perceived barriers to the wider uptake of anticoagulation, quality control of anticoagulation and their assessment of strategies to assist in managing NVAF.

Results

596 out of 924 eligible GPs responded (64.4% response rate). The majority of GPs recognised that the benefits of warfarin outweighed the risks for three case scenarios in which warfarin is recommended according to Australian guidelines. In response to a hypothetical case scenario describing a patient with a supratherapeutic INR level of 5, 41.4% of the 596 GPs (n = 247) and 22.0% (n = 131) would be "highly likely" or "likely", respectively, to cease warfarin therapy and resume at a lower dose when INR levels are within therapeutic range. Only 27.9% (n = 166/596) would reassess the patient's INR levels within one day of recording the supratherapeutic INR. Patient contraindications to warfarin was reported to "usually" or "always" apply to the patients of 40.6% (n = 242/596) of GPs when considering whether or not to prescribe warfarin. Patient refusal to take warfarin "usually" or "always" applied to the patients of 22.3% (n = 133/596) of GPs. When asked to indicate the usefulness of strategies to assist in managing NVAF, the majority of GPs (89.1%, n = 531/596) reported that they would find patient educational resources outlining the benefits and risks of available treatments "quite useful" or "very useful". Just under two-thirds (65.2%; n = 389/596) reported that they would find point of care INR testing "quite" or "very" useful. An outreach specialist service and training to enable GPs to practice stroke medicine as a special interest were also considered to be "quite" or "very useful" by 61.9% (n = 369/596) GPs.

Conclusion

This survey identified gaps, based on GP self-report, in the current care of NVAF. GPs themselves have provided guidance on the selection of implementation strategies to bridge these gaps. These results may inform future initiatives designed to reduce the risk of fatal and disabling stroke in NVAF.

Use of computer decision support interventions to improve medication prescribing in older adults: a systematic review

BACKGROUND

Older adults take multiple medications and are at high risk for adverse drug effects.

OBJECTIVE

This systematic review was conducted to describe the impact of computer decision support (CDS) interventions designed to improve the quality of medication prescribing in older adults.

METHODS

PubMed and EMBASE databases were searched from January 1980 through July 2007 (English-language only); studies were eligible if they described a CDS intervention intended to improve medication prescribing in adults aged > or =60 years. Studies were retained if they were observational or experimental in design and reported > or =1 process or clinical outcome measurement related to medication prescribing. In the main analysis, study characteristics and major outcome results were extracted. A combination of searches was performed using relevant medical subject headings: aged; drug therapy, computer-assisted; medication errors; medication errors/prevention and control; decision making, computer-assisted; decision support systems, clinical; and clinical pharmacy information systems.

RESULTS

After review of study abstracts, 10 articles met the eligibility criteria. Of those 10 studies testing CDS interventions, 8 showed at least modest improvements (median number needed to treat, 33) in prescribing, as measured by minimizing drugs to avoid, optimizing drug dosage, or more generally improving prescribing choices in older adults (according to each study's intervention protocols). Findings for the impact of CDS interventions on clinical outcomes were mixed and were reported for only 2 studies.

CONCLUSIONS

Various types of CDS interventions may be effective in improving medication prescribing in older adults, but few studies reported clinical outcomes related to changes in medication prescribing. Data from this study should help to guide refinement and testing of future CDS interventions that specifically target older adult populations that are taking multiple medications.

Qualitative methods in a randomised controlled trial: the role of an integrated qualitative process evaluation in providing evidence to discontinue the intervention in one arm of a trial of a decision support tool

OBJECTIVE

To understand participants' experiences and understandings of the interventions in the trial of a computerised decision support tool in patients with atrial fibrillation being considered for anti-coagulation treatment.

DESIGN

Qualitative process evaluation carried out alongside the trial: non-participant observation and semistructured interviews.

PARTICIPANTS

30 participants aged >60 years taking part in the trial of a computerised decision support tool.

RESULTS

Qualitative evidence provided the rationale to undertake a decision to discontinue one arm of the trial on the basis that the intervention in that arm, a standard gamble values elicitation exercise was causing confusion and was unlikely to produce valid data on participant values.

CONCLUSIONS

Qualitative methods used alongside a trial allow an understanding of the process and progress of a trial, and provide evidence to intervene in the trial if necessary, including evidence for the rationale to discontinue an intervention arm of the trial.

Medical communication and technology: a video-based process study of the use of decision aids in primary care consultations

BACKGROUND

Much of the research on decision-making in health care has focused on consultation outcomes. Less is known about the process by which clinicians and patients come to a treatment decision. This study aimed to quantitatively describe the behaviour shown by doctors and patients during primary care consultations when three types of decision aids were used to promote treatment decision-making in a randomised controlled trial.

METHODS

A video-based study set in an efficacy trial which compared the use of paper-based guidelines (control) with two forms of computer-based decision aids (implicit and explicit versions of DARTS II). Treatment decision concerned warfarin anti-coagulation to reduce the risk of stroke in older patients with atrial fibrillation. Twenty nine consultations were video-recorded. A ten-minute 'slice' of the consultation was sampled for detailed content analysis using existing interaction analysis protocols for verbal behaviour and ethological techniques for non-verbal behaviour.

RESULTS

Median consultation times (quartiles) differed significantly depending on the technology used. Paper-based guidelines took 21 (19-26) minutes to work through compared to 31 (16-41) minutes for the implicit tool; and 44 (39-55) minutes for the explicit tool. In the ten minutes immediately preceding the decision point, GPs dominated the conversation, accounting for 64% (58-66%) of all utterances and this trend was similar across all three arms of the trial. Information-giving was the most frequent activity for both GPs and patients, although GPs did this at twice the rate compared to patients and at higher rates in consultations involving computerised decision aids. GPs' language was highly technically focused and just 7% of their conversation was socio-emotional in content; this was half the socio-emotional content shown by patients (15%). However, frequent head nodding and a close mirroring in the direction of eye-gaze suggested that both parties were active participants in the conversation

CONCLUSION

Irrespective of the arm of the trial, both patients' and GPs' behaviour showed that they were reciprocally engaged in these consultations. However, even in consultations aimed at promoting shared decision-making, GPs' were verbally dominant, and they worked primarily as information providers for patients. In addition, computer-based decision aids significantly prolonged the consultations, particularly the later phases. These data suggest that decision aids may not lead to more 'sharing' in treatment decision-making and that, in their current form, they may take too long to negotiate for use in routine primary care.