Personalized Medicine for Chronic, Complex Diseases: Chronic Obstructive Pulmonary Disease as an Example

Josiah E Radder; Steven D Shapiro; Annerose Berndt


Personalized Medicine. 2014;11(7):669-679. 

In This Article

Bringing Personalized Medicine for Chronic Disease to the Bedside

As our grasp of the complexity of chronic diseases strengthens, the ultimate goal of personalized medicine will be to once again simplify that understanding into clinically actionable results. Guidelines such as those laid out by GOLD for COPD have been instrumental in standardizing care and implementing phenotype-based decision-making based on measurable characteristics, such as exacerbation history and pulmonary function.[83] However, they take only a small amount of information about a patient's phenotype into account. At the same time, they have been criticized as being complex to implement broadly, particularly in primary care settings, where studies from many regions of the world have confirmed that adherence to guidelines is inconsistent.[84–86] Will methods for personalizing medicine for chronic disease be discovered but fail in implementation due to their complexity?

Fortunately, we believe that improvements in health information technology will not only aid in discovering personalized therapies, but will also help physicians to employ them at the bedside. Clinical decision-support systems (CDSSs), most of which currently exist within the EHR, offer passive or active prompts to clinicians based on the integration of individual patient information with a knowledge base (e.g., recommended guidelines) in order to improve evidence-based practice and improve patient care. An early meta-analysis of the effectiveness of CDSSs demonstrated that these tools can be effective in improving patient care, particularly when including automatic notifications via computer that offer recommendations, not assessments, at the time of decision-making.[87] Recent studies have demonstrated that these tools can also improve evidence-based care by applying algorithm-based support even for complicated guidelines.[88]

Several changes in CDSSs will be necessary in order to properly accommodate chronic disease. Although CDSSs in chronic diseases such as asthma have been shown to be effective, they tend to be underutilized, at least in part because their advice is viewed by providers as neither pertinent nor timely.[89] The pertinence of support to clinical practice can be improved by improving the context dependence of the results.[90] Context will best be improved by improving phenotyping, as previously described, but should also take into account factors that are external to the disease (e.g., whether a patient's insurance covers a proposed drug).

Finally, as CDSSs become more common and offer support for more complex decisions, it will be essential that physicians and professional organizations remain closely involved in both their design and maintenance. Algorithms designed to offer support on guidelines in CDSSs must be rapidly updated upon changes to those guidelines. Who will do this and how it will be instituted remains to be determined. Moreover, the tradeoff between mistakes made by CDSSs on an individual level and improved outcomes at a population level must be considered before CDSSs are put into widespread use.[91]