{
  "abstract": "Objectives Decision support tools and electronic alerts could improve clinical care; but the effect of cognitive biases that could influence the effectiveness of these tools has been unclear. Redelemeir and Shafir (1995) found that presenting multiple alternatives increased the likelihood that physicians would remain with an existing management plan, even if it could be considered low-value. This study concluded that clinicians experience ‘status quo bias’ when presented with multiple alternatives and thus we should avoid presenting clinicians with too many options. This hallmark study has been used to inform many quality improvement projects. The original finding has not been replicated.We investigated whether presenting two or more high-value treatment alternatives vs one high-value treatment alternative could better shift primary care physicians toward high-value care.Methods This was a randomised vignette experiment. Primary care physicians completed two clinical scenarios; one on opioid prescribing for chronic back pain and one on surgery for knee osteoarthritis. In the control group, physicians received one high-value treatment alternative. In the intervention group physicians could receive two, three, or four high-value alternatives. The primary outcome was the proportion of physicians that chose a high-value treatment alternative. Analysis used generalized estimating equation (GEE) to estimate the odds of the intervention on the probability of clinicians choosing an alternative.Results We recruited 402 practising primary care physicians from the United States. Physicians were more likely to choose a high-value treatment option when presented with two or more alternatives to the existing management plan compared to only one alternative (44% in control versus 62% in intervention; adjusted OR=1.90 (1.09 to 3.30)). The intervention had larger effects in the scenario about opioids (31% versus 56%, unadjusted OR=2.95 (1.96 to 4.45) than the scenario about surgery (58% versus 68%, unadjusted OR=1.56 (1.04 to 2.34)). Exploratory subgroup analysis found no difference when physicians were presented with two, three or four alternatives.Conclusions Offering a clinician two or more high-value alternatives to an existing management plan, for example using decision support or a computer alert, could be effective at improving care. Our data challenge claims that physicians experience status-quo bias in medical decision-making situations that offer multiple alternatives. Designers of decision support tools aiming to reduce low value care should consider including several high value alternatives.",
  "authors": [
    {
      "affiliations": [
        "University of Sydney, Sydney, Australia"
      ],
      "name": "Gemma Altinger"
    },
    {
      "affiliations": [
        "University of Sydney, Sydney, Australia"
      ],
      "name": "Chris G Maher"
    },
    {
      "affiliations": [
        "University of Sydney, Sydney, Australia"
      ],
      "name": "Catilin Jones"
    },
    {
      "affiliations": [
        "University of Technology Sydney, Sydney, Australia"
      ],
      "name": "Jason Collins"
    },
    {
      "affiliations": [
        "Northwestern University Feinberg School of Medicine, Chicago, USA"
      ],
      "name": "Jeffrey A Linder"
    },
    {
      "affiliations": [
        "University of Sydney, Sydney, Australia"
      ],
      "name": "Katy Bell"
    },
    {
      "affiliations": [
        "University of Sydney, Sydney, Australia"
      ],
      "name": "Chung-Wei Christine Lin"
    },
    {
      "affiliations": [
        "University of Sydney, Sydney, Australia"
      ],
      "name": "Marguerite Tracy"
    },
    {
      "affiliations": [
        "University of Sydney, Sydney, Australia"
      ],
      "name": "Farzaneh Boroumand"
    },
    {
      "affiliations": [
        "University of Sydney, Sydney, Australia"
      ],
      "name": "Adrian C Traeger"
    }
  ],
  "title": "063 Decision support is more effective in shifting primary care physicians towards high-value treatment for back pain when more than one high value alternative is offered: a randomised experiment",
  "uid": "d0b4ec6f-a287-5eea-91d5-b6dd3147cc9a"
}