Use of an on-line patient portal in a depression collaborative care management program

•We reviewed use of a patient portal to connect with depression care managers.•Portal users were more likely to be younger, married and female.•Portal users were more likely to complete six month depression questionnaires.•Portal users had higher rates of remission at six months in intention to treat analysis.•This difference disappeared when comparing only those who completed 6 month PHQ-9's.

BackgroundEvidence for patient portals on health outcomes is still accumulating.MethodsWe performed a retrospective study of patients in a depression collaborative care management (CCM) program comparing dropout rates and outcomes at six months in patients who used the portal with their CCM managers versus those who did not use the portal for this purpose.Results1769 patients were analyzed, of which 272 (15.4%) used the portal to connect with their CCM managers. Portal users were more likely to be younger (mean age 37.8 years versus 42.6 years p<0.001), married (57.5% versus 48.2% p=0.003) and female (81% versus 69.4% p<0.001) compared to non-portal users. Completion of six month PHQ 9 forms was significantly higher in the portal group (75.7% vs. 65.7%, p=0.001) with an adjusted odds ratio (AOR) of 1.68 (1.25, 2.30, p<0.001). Using an intention to treat analysis, remission at 6 months was significantly higher in the portal group (45.2% vs 38.5%, p=0.038) with an AOR of 1.32 (1.00-1.73, p=0.47). However analysis of only those who completed a PHQ-9 at six months showed no difference in depression remission (59.7% vs 58.7%, p=0.78) with an AOR of 1.02 (0.75, 1.42. p=0.87). There was no difference in dropout rates between portal users compared to non-portal users (31.6% vs. 31.7%, p=0.97) with an AOR of 0.95 (0.71, 1.27, p=0.69).LimitationsLimitations include the observational study design and a predominantly white study population.ConclusionUse of a patient portal in depression CCM allows additional options for patients to participate without negatively influencing clinical outcomes.