Article Text
PDF
Statistics from Altmetric.com
Question
Question: How effective is a practice-based programme for screening, diagnosing and managing depression in postpartum mothers?
Patients: A total of 2343 women between 5 and 12 weeks postpartum from 28 eligible primary care practices. The women had to be at least 18 years, speak English or Spanish and have been receiving continuing care at the family medical practice where they were enrolled. Eligible practices had to have provided maternity or well-baby care services to more than 30 individuals in the year prior to the study and not have been routinely screening for postpartum depression.
Setting: Twenty-eight family medicine practices, USA; March 2006–August 2010.
Intervention: A practice-wide intervention consisting of postpartum depression screening, diagnosis, evaluation and management. Intervention sites received training on postpartum depression screening and diagnosis using the Edinburgh Postnatal Depression Scale (EPDS) and the nine-item Patient Health Questionnaire PHQ-9, as well as training and practice of nursing telephone calls. Intervention practices were given tools containing an outline for follow-up visits and written materials for nursing follow-up phone calls related to medication and side effects. Usual care (control) sites continued to provide the same postpartum care as before the study. Patients from intervention and usual care practices completed two self-report questionnaires (EPDS and PHQ-9) at baseline, 6 and 12 months.
Outcomes: Primary outcome: A reduction in PHQ-9 score ≥5 points at 6 or 12 months postpartum was considered indicative of a clinical improvement or response to therapy. A PHQ-9 score of 10 or greater was considered indicative of postpartum depression if no other cause for the depressive symptoms were found by the physician. In cases where the EPDS score was greater than 10 (indicating possible depression) and the PHQ-9 was not elevated, the physician used clinical judgement.
Patient follow-up: Questionnaires were completed by 77% of eligible participants at baseline, 68% at 6 months and 62% at 12 months.
Methods
Design: Cluster randomised controlled trial (with randomisation at the practice level).
Allocation: Unclear.
Blinding: Unclear if outcome assessors were blinded; due to nature of intervention, practices could not be blinded.
Follow-up period: 12 months.
Main results
At baseline, elevated screening scores indicating depression were reported for 34.5% of women included in the analysis, with rates similar across intervention and usual care groups. Diagnosis, therapy initiation and referral for psychiatric care were significantly more likely among patients from the intervention practices. Among participants with elevated baseline depression scores, significantly more women in the intervention group had a five point or greater drop in PHQ-9 scores at 12 months compared with the control group (adjusted OR 1.74, 95% CI 1.05 to 2.86). There was no significant difference between the groups at 6 months in terms of clinical improvement among women diagnosed with postpartum depression.
Conclusions
A primary care-based intervention for screening, diagnosing and management improves self-reported depression in postpartum mothers.
Notes: After 24 months, the usual care sites were crossed over to the intervention. The authors report that outcomes included ‘women enrolled after the usual care sites crossed over’. It is not clear how many women crossed over. Baseline characteristics differed between groups with intervention participants less educated, poorer and less likely to be married.
Abstracted from
Commentary
Postpartum depression (PPD) is a major public health concern that affects between 7 and 15% of childbearing women. PPD has a negative impact not only on mothers’ mental and physical health, but also on mother–infant interaction and infants’ neurodevelopment. A number of clinical interventions that investigated screening and standardised treatment of PPD in obstetric and general practice settings failed to find significant improvements in mothers’ health outcomes. However, with a few exceptions, most of these previous studies had major methodological flaws in their research design or lacked statistical power due to small sample sizes or high dropout rates. As a result, recommendations for universal perinatal screening are still subject to controversy. In a recent study from Yawn and colleagues, 1897 women between 5 and 12 weeks postpartum who were receiving continuing care in 28 primary care teams in the USA were randomly allocated to intervention or treatment as usual. In this study, randomisation occurred at the ‘practice’ level with a crossover design, and the intervention was based on systematic screening, assessment and treatment of PPD. Results showed that women allocated to the primary care-based intervention were more likely to receive diagnosis and treatment for PPD, and were more likely to improve from PPD at 6 (p=0.07) and 12 months (p=0.001) of follow-up. Moreover, the intervention and being diagnosed with PPD were independent predictors of clinical improvement. It is worth noting that only half of women in the intervention group received medication and, even more surprisingly, only 20% of these women received counselling. These numbers raise an issue of potential limited access to optimal treatment for PPD, which may be due, in part, to the fact that almost one-third of women in the intervention group lost their medical insurance at 2 months postpartum. Thus, these findings may not be generalised to countries with primary public health systems. Nevertheless, the TRIPPD study describes an effective approach that is feasible to implement in primary care centres, which suggests that it is time to rethink effectiveness in screening and treatment of PPD.
Footnotes
Correspondence to Barbara Yawn, Research Department Olmsted Medical Centre 210 Ninth St SE Rochester, MN 55904, USA; byawn@olmmed.org
Sources of funding The Agency for Healthcare Research and Quality.