Managing Postpartum Depression: Provider Opinions and Practices
Despite the availability of effective screening tools for postpartum depression, the condition remains largely underdiagnosed in the United States
By Joanna Drowos, DO, MPH
Postpartum depression is the most common complication of childbearing, with an estimated 10-20 percent of women in the United States experiencing symptoms within six months of delivery.1
Recent media attention has focused on the potential negative impact upon family members of women experiencing postpartum depression.2 Of particular concern are the cognitive, social and behavioral consequences on the development of young children in these situations.3 Such concerns have led to recommendations for the reduction of postpartum depression as a postnatal complication.
Previous studies confirm that detection of postpartum depression is significantly higher with the use of a screening instrument during routine clinical evaluation. Various healthcare provider specialty groups have been identified as having the potential to play a larger role in the early identification of affected women, especially obstetricians, pediatricians, and family physicians. In prior survey responses, specialty groups revealed that screening is not universal and the use of postpartum depression screening tools is uncommon.4
Barriers to screening have been identified, and include the continued stigma associated with discussing mental illness, insufficient time for adequate history, insufficient training or knowledge to diagnose, counsel or treat patients, and even that mothers looking for help may not consider their child’s physician as a source of help.3
Performing an analysis of all providers in a community of clinics addressing the healthcare needs of new mothers and children can identify trends and opinions towards screening for postpartum depression. Surveying guidelines and practices across specialties within a community can provide a comprehensive look at perceptions of postpartum depression within that community, and allows for designing interventions through the local Health Department to improve an entire community’s practices.
Methods
Sample Characteristics
In our analysis, all health care providers providing services to new mothers and newborns at the test survey site Health Department clinic locations were eligible for survey participation. At the four test survey site Health Department clinics providing child and maternal services, there are 50 practitioners involved including 19 obstetricians, 12 pediatricians, and 19 general medicine providers. This group of practitioners includes a combination of physician assistants, nurse practitioners, and physicians.
Procedure
Following IRB (Institutional Review Board) approval from both the Florida Department of Health and Nova Southeastern University, a cross-sectional analysis was performed based on a 21-item questionnaire. Questions were based on prior studies evaluating attitudes and practices in screening for postpartum depression. The questionnaire was given to Health Center Directors to distribute to the healthcare providers in each location, by both email and standard mail. Return envelopes were provided as well as an email address for responses.
Only the principal investigator had access to the responses and entered all data into a chart without identifying responses by subject. Epi Info™ was used for data entry to maintain consistency between responders and for easy conversion to a spreadsheet. Epi Info™ is a public domain software package designed for the global community of public health practitioners and researchers. It provides for easy form and database construction, data entry, and analysis with epidemiological statistics, maps, and graphs. (For more information and/or download, visit cdc.gov/epiinfo/installation.htm)
SAS (Statistical Analysis System) software was used to examine relationships between the 34 variables included in the questionnaire. Chi-square values and p-values were calculated for measuring the significance of relationships between variables.
Results
Sample Demographics
Twenty-four providers completed and returned surveys for consideration, a 50 percent response rate. Of the providers who responded, the majority was female (62.5%) and more allopathic physicians responded than any other type of provider (54.17%.)
A larger proportion of providers practicing in the field of Ob-Gyn responded than from any other specialty (29.17%). The majority of providers attended professional school in the United States (66.66%) and were board certified in their particular specialty (83.3%.)
Time in practice for providers ranged from one to 32 years with an average of 16.42 years in practice, while the number of visits with new mothers/infants per year ranged from 0 to 4,800, with an average of 803.71 visits per provider. The majority of respondents were career services employees (62.5%.)
Compared with the known characteristics of our total population of providers, pediatricians were largely underrepresented. In the total clinic provider population, 12 of the 51 practitioners work in pediatrics (23.5%), while 19 of the 51 providers work in Ob-Gyn (37.2%.) The remaining providers work in general medicine, 11 (21.5%) as internists and eight (15.7%) as family physicians.
Females were also over represented in the sample since 24 (47.1%) providers throughout the clinics are female and the remaining 27 (52.9%) are male. Respondents were more likely to be female which is not true in the total provider population. Allopathic physicians were also slightly underrepresented, since there are actually 10 ARNPs (19.6%), three PA-Cs (5.9%), six osteopathic physicians (11.8%), and 32 allopathic physicians (62.7%). A larger number of MD respondents would have made the sample more representative of the total provider population.
Subjects were also asked whether they had received any training to aid their understanding of postpartum depression through four different methods available for educating providers. A total of seven out of the 24 subjects (29.2%) responded they had never received training through any of the means listed.
As far as training in professional school, seven respondents (45.8%) reported that they had covered the subject of postpartum depression. Only six subjects (25%) responded they had received training during residency. Only 11 subjects (45.8%) reported accessing training through the medical literature, while 12 (50%) reported participating in postpartum depression training through continuing education activities.
When asked how they felt about the statement “Postpartum depression is common enough to warrant screening,” 20 providers (83.3%) agreed or strongly agreed. One provider disagreed with the statement “Postpartum depression is serious enough to warrant screening,” while 21 others (87.5%) agreed or strongly agreed with this statement. When asked about the statement, “Screening for postpartum depression is effective for identifying affected women,” one provider disagreed, while 15 (62.5%) agreed or strongly agreed.
For the statements “Postpartum depression affects the children of affected mothers” and “Postpartum depression affects the spouses of affected women,” one provider strongly disagreed with each of these statements. Nineteen providers (79.2%) agreed or strongly agreed with the first statement and 18 (75%) agreed or strongly agreed with the second statement. Fifteen providers (62.5%) either agreed or strongly agreed with the statement “Therapy for postpartum depression is effective.”
Personal Attitudes
Providers were also asked to indicate their level of agreement with eight statements regarding personal opinions about postpartum depression. (Table 3) When asked about agreement with the statement “I possess adequate knowledge about postpartum depression,” nine providers (37.5%) either disagreed or strongly disagreed. Eleven providers (45.8%) disagreed or strongly disagreed with the statement “I am aware of formal screening methods for postpartum depression” while only four providers (16.7%) agreed or strongly agreed with the statement “I feel comfortable providing treatment for postpartum depression.” Only eight providers (33.3%) agreed or strongly agreed with the statement “I possess adequate knowledge of referral resources for postpartum depression.”
Subjects were also asked separately about their feelings about whether screening at every well-child and postpartum visit would take too much time, and whether screening at each of these visits would be effective. When given the statement “Screening for postpartum depression at every well-child visit up to one-year of age would not be effective” and “Screening for postpartum depression at each postpartum visit would not be effective” the majority of providers had no opinion – 11 (45.8%) and 10 (41.7%) respectively.
The majority of providers (13 or 54.2%) also disagreed or strongly disagreed with both statements: “Screening for postpartum depression at every well-child visit up to one-year of age would not be effective” and “Screening for postpartum depression at each postpartum visit would not be effective.”
Practices
Providers were asked about the frequency with which they perform seven specific screening management activities. (Table 4) Many of the statements ask specifically about well-child and postpartum visits in which many of the providers are not involved. Of the 10 providers who applied to the statement “New mothers are formally screened for postpartum depression during well-child visits during the first-year of life” four reported screening rarely or never.
When asked whether “New mothers are informally screened for postpartum depression during well-child visits during the child’s first-year of life,” 15 providers indicated this statement was not applicable to them, and five of the remaining nine providers reported always or often performing screening.
Eleven of the 13 providers reported, “New mothers are formally screened for postpartum depression during postpartum visits” while eight of 12 providers, reported that “New mothers are informally screened for postpartum depression during postpartum visits” always or often. In terms of whether “New mothers are provided treatment for symptoms of postpartum depression” five of the 10 providers who felt this question applied to them reported new mothers are rarely provided treatment.
Ten of the 15 providers who care for new mothers reported that rarely or never, “New mothers are referred for further evaluation of postpartum depression symptoms.” In terms of providing treatment, 10 of the 14 providers who felt the statement “New mothers are referred for treatment of postpartum depression symptoms” applied to them, reported that they refer new mothers rarely.
Relationships
SAS software was used to create tables and calculate chi square values and p-values to determine whether significant relationships exist between certain variables. Analysis was performed comparing type of degree and likelihood of receiving training, gender, and general beliefs regarding postpartum depression, personal attitudes, and specialty and specialty practice with number of visits with infants/new mothers per year. For the majority of relationships examined, statistical significance was not reached.
Comparing type of provider degree and the likelihood of receiving training through residency revealed a significant relationship (chi-square value 7.6923, DF 3, p-value 0.05). None of the six ARNPs received training, as well as the one PA who responded. Three of the four osteopathic physicians who responded did receive training compared with only three of the 14 allopathic physicians who answered that they received training during residency.
Comparing general beliefs about postpartum depression with gender revealed a significant relationship (chi-square value 7.4785, DF 3, p-value 0.05.) Thirteen of the 15 females who responded to the survey stated that they agreed or strongly agreed with the statement that “postpartum depression affects the children of affected mothers” compared with only six of the nine males who replied similarly.
Comparing personal attitudes by specialty practiced revealed the following significant relationship (chi-square value 23.9016, DF 12, p-value 0.02.)
When asked about agreement with the statement “I possess adequate knowledge about postpartum depression” Ob-Gyns were most likely to agree, while internists and preventative medicine practitioners were most likely to disagree.
Another significant relationship was found by comparing specialty practiced and personal attitudes (chi-square 31.0419, DF 16, p-value 0.01.) When asked about agreement with the statement, “Screening for postpartum depression at each postpartum visit takes too much time” three of the seven Ob-Gyns who replied agreed, while the other four disagreed. The majority of the family physicians, who replied, disagreed, while the internists and preventive medicine practitioners had no opinion.
Another significant relationship was found by comparing specialty practiced and personal attitudes (chi-square 22.5483, DF 12, p-value 0.03.) When asked about agreement with the statement “Screening for postpartum depression at each postpartum visit would not be effective” most of the Ob-Gyns disagreed, while the majority of other specialties did not have an opinion.
When comparing specialty practiced with number of visits per year with infants/new mothers a significant relationship was found (chi-square 17.2286, DF 8, p-value 0.02.) The number of visits was classified into three groups:
1. Range 0
2. Range 1 – 500
3. Range 501+
The majority of Ob-Gyns and pediatricians provided the most visits, while all of the other specialties provided fewer visits.
Discussion
Our study results were consistent with prior studies examining trends in physician screening and management of postpartum depression.
A study by Seehusen’s Group examined the rates at which family physicians were screening for postpartum depression. Results indicated that family physicians believe that postpartum depression is serious, identifiable, and treatable; however screening is not universal, and use of specific screening tools is uncommon.
A study by Curie’s group examined the role that pediatricians could play in screening and intervening in cases of postpartum depression.
Only 57 percent of the pediatricians who completed surveys responded that they felt it was their responsibility to recognize the problem, and only 48 percent conducted additional assessment once a problem was recognized.
Major barriers to diagnosis included insufficient time for adequate history, and inadequate training or knowledge to diagnose, counsel, or treat postpartum depression.
This study characterized provider attitudes and practices across specialties and provider types by drawing on the health care providers employed by the Survey site Health Department. This limited sample size could have impacted the statistical significance found between sets of variables.
For example, the relationship between type of degree and whether training occurred revealed a test of independence showed the two variables as insignificantly associated; however, the p-value of 0.0528 was considered a significant p-value because of the sample size.
Some reclassification of variables was performed during data analysis for simplification. The number of visits with infants/new mothers was broken down into three groups, even though providers had been asked to write in a number rather than select from a choice of ranges. This also occurred with specialty selection, as some providers indicated they were certified in a specialty that was not practiced within our clinical setting, and thus these providers were reclassified to the specialty they participated in that was relevant to the study.
Time also proved to be a limitation in completing this study, as the IRB approval process took longer than planned and cut into data collection time. Initially the study design called for a two-month period of data collection, however this was limited due to receiving late approval and various other scheduling conflicts. The 50 percent response rate could have been higher providing a larger sample size had more time been devoted to the data collection process.
A follow-up study with a larger sample size could better establish these trends and relationships, although even with preliminary data there are opportunities to improve the standard of care provided in the test survey site Health Department Clinics.
Opportunities to Improve
Instituting a policy to promote screening by all providers who come in contact with children and new mothers could increase awareness of the problem and encourage better compliance with screening and referral regimens. More importantly, additional provider training could improve their knowledge base and shape some of the assumptions they have regarding postpartum depression, to improve care for clinic patients.
To address the deficiency in knowledge, the Health Department could create a public health seminar to educate all providers about postpartum depression. This could also be the venue to introduce new clinical policies to promote screening and treatment.
A committee could be created to oversee the development of policy, consisting of professionals from pediatrics, maternity, Healthy Start, community health and administration. First, the committee could select a screening tool to use universally in the clinics, such as the Edinburgh Depression scale, which has proven effective in studies.2
Second, the committee could also develop treatment protocols, beginning with creating educational materials for patients about the disease, resources to get help, as well as for the medication management where indicated.
All of these materials would be widely available in the clinics for easy use. This study suggests that there are differences across provider types and characteristics, and that there are perceived gaps in knowledge pertaining to the management of postpartum depression.
Click here to view Tables and Graphs
Dr. Drowos is a 2004 graduate of Nova Southeastern University College of Osteopathic Medicine. She earned her Masters of Public Health there in 2003. She completed her residency in Preventive Medicine at the Palm Beach County Health Department in 2006 and has earned certification in General Preventive Medicine/Public Health. She is finishing her final year of family medicine at Broward General Medical Center in Ft. Lauderdale, Florida.
References