Using Algorithms and Computerized Decision Support Systems to Treat Major Depression

Richard C. Shelton, MD (Chair)

Department of Psychiatry, Vanderbilt School of Medicine, Nashville, Tennessee

Madhukar H. Trivedi, MD

Mood Disorders Research Program and Clinic, University of Texas Southwestern Medical Center, Dallas

Measurement-Based Care and Algorithms

In November 2010, the APA released updated treatment guidelines1 for MDD that stress the use of measurement-based care and treatment algorithms. In measurement-based care, clinicians measure symptom severity and side effect burden (by using tools such as the QIDS-SR-16 and FIBSER) and assess patients’ medication adherence, functioning, and suicide risk.

Information gathered from initial assessments and from consistent patient follow-ups provides feedback for clinicians to use at critical decision points, which are based on a state-of-the-art algorithm such as the TMAP treatment algorithm for nonpsychotic MDD. The TMAP algorithm offers multiple treatment options based on the patient’s clinical status at each decision point, and treatments with the best evidence and the best risk/benefit ratios are presented in the initial steps. The use of the TMAP treatment algorithm has been shown to provide greater symptom reduction than treatment as usual.2 However, implementation of algorithms and guidelines in clinical practice has been inadequate to date.3

A variety of barriers prevent the implementation of algorithms in real-world treatment settings. For example, physicians cite a lack of awareness, lack of time, and insufficient staff support as barriers to adhering to guidelines.4 Physicians’ acceptance of an algorithm is also influenced by whether the algorithm offers treatment alternatives, ie, whether they view it as being flexible enough to manage diverse patients.5 Ease of use and accessibility are also factors in sustaining adherence to a treatment algorithm. These barriers have created a need for an effective means of integrating evidence-based care into clinical practice. One approach to improve the implementation of guidelines and algorithms is the development of computerized decision support systems (CDSS).

Computerized Decision Support Systems

A CDSS can make measurement-based care strategies accessible and user-friendly for physicians and their staffs, individualize treatment options according to each patient’s circumstances, and provide evidence-based guideline information at the point of care when decisions about treatment are being made. To investigate how best to integrate MDD treatment algorithms into clinical practice, Trivedi and colleagues3 developed CompTMAP, a CDSS that combines the core components of measurement-based care and an adapted version of the TMAP algorithm.

 

CompTMAP is a software program that mimics a clinical visit: the clinician begins with the patient selection screen, then moves to the diagnosis screen (for new patients) or the evaluation screen (for returning patients), and then proceeds to the treatment selection screen (AV 1AV 1). In each screen, information is recorded, stored, and displayed graphically, while progress notes are automatically generated. Data collected from the patient through patient-rated scales and questionnaires can be entered directly into the system, thereby combining patient-level and clinician-level information into a single record.

The patient’s information from the evaluation screen, including the type, duration, and dosage of medications, along with the patient’s medication adherence, response, and side effect burden, is filtered through the treatment algorithm to help the physician analyze the patient’s current clinical status. The application offers treatment options, provides suggestions for medications to treat the diagnosis as well as associated symptoms and side effects, and provides safety-related reminders and screen prompts that assist the physician with other treatment considerations (eg, blood tests that are required with certain medications). Throughout this process, the physician may override the system’s treatment suggestions and choose other interventions for the patient.

More

In addition, the CDSS prompts the clinician to schedule follow-up visits according to assessments of the patient’s symptoms, treatment response, adverse effects, and adherence; patients who experience less than adequate improvement with treatment will be seen more often than treatment responders. The measurement-based care in the CDSS also allows patients to engage in their own treatment in that, if they are measuring their symptoms at home and the symptoms are worsening or are not improving, patients can contact the clinician’s office to schedule a visit.

 

In a proof-of-concept study6 evaluating the feasibility and effectiveness of implementing the CDSS in primary care practices, patients with depression treated according to the CDSS had significantly greater reduction in symptom severity than those receiving treatment as usual (AV 2AV 2). The CDSS group had a higher number of treatment visits than the usual care group, which was the only treatment difference between the groups that was significantly different.

Barriers to using a CDSS include physicians’ computer literacy levels and facility hardware/software capabilities; additionally, some duplication exists with offices transitioning from paper to EHRs. A study7 is currently being conducted to test a method of instituting measurement-based care in clinical practice by merging the CompTMAP CDSS with an existing EHR system so that physicians will have only 1 software system to navigate. When concluded, the study will include an evaluation of the impact of the CDSS/EHR system on physicians and their patients and its successfulness in promoting measurement-based care into clinical practice.

Conclusion

A CDSS helps clinicians optimize treatment by implementing measurement-based care through the continued assessment of depressive symptoms, treatment adherence, medication-induced adverse events, functioning, and suicide risk. The CDSS is a resource that provides readily accessible information to clinicians during the decision-making process, in addition to safety alerts and treatment prompts. New guidelines can be incorporated into the CDSS as soon as they are available. As more information becomes available regarding factors that affect the treatment of depression—including the results from an increasing number of genetic, neurobiologic, biologic, and treatment moderator studies—having a matrix in which to distill this information will help to guide clinicians on treatment choices for individual patients. The major goal is for clinicians to increase the number of patients who recover from depression in a reasonable period of time and remain symptom-free.

For Clinical Use

  • Follow treatment algorithms to improve patient outcomes
  • Use measurement-based care to systematically assess patients’ depressive symptoms, treatment adherence, functioning, side effect burden, and suicide risk
  • Monitor patients regularly, and assess treatment options at critical decision points
  • Consider EHRs and a CDSS to facilitate measurement-based care

Abbreviations

APA=American Psychiatric Association, CDSS=computerized decision support system, EHR=electronic health record, FIBSER=Frequency, Intensity, and Burden of Side Effects Ratings, HDRS=Hamilton Depression Rating Scale, MDD=major depressive disorder, QIDS-SR-16=16-item Quick Inventory of Depressive Symptomatology–Self-Rated, TMAP=Texas Medication Algorithm Project

Take the online posttest.

References

  1. American Psychiatric Association. Practice Guideline for the Treatment of Patients With Major Depressive Disorder, Third Edition. Washington, DC: American Psychiatric Association; 2010. http://www.psychiatryonline.com/pracGuide/pracGuideTopic_7.aspx. Accessed January 19, 2011.
  2. Trivedi MH, Rush AJ, Crismon ML, et al. Clinical results for patients with major depressive disorder in the Texas Medication Algorithm Project. Arch Gen Psychiatry. 2004;61(7):669–680.
  3. Trivedi MH, Kern JK, Grannemann MS, et al. A computerized clinical decision support system as a means of implementing depression guidelines. Psychiatr Serv. 2004;55(8):879–885.
  4. Smith L, Walker A, Gilhooly K. Clinical guidelines of depression: a qualitative study of GPs' views. J Fam Pract. 2004;53(7):556–561.
  5. Trivedi MH. Using treatment algorithms to bring patients to remission. J Clin Psychiatry. 2003;64(suppl 2):8–13.
  6. Kurian BT, Trivedi MH, Grannemann BD, et al. A computerized decision support system for depression in primary care. Prim Care Companion J Clin Psychiatry. 2009;11(4):140–146.
  7. Trivedi MH, Daly EJ, Kern JK, et al. Barriers to implementation of a computerized decision support system for depression: an observational report on lessons learned in "real world" clinical settings. BMC Med Inform Decis Mak. 2009;9:6.
​​​​​​​​​