Discussion
One in three postnatal women had a high EPDS-4 score, which is three times higher than ten years earlier. Our findings reveal considerable dissatisfaction with the care in the maternity ward among mothers during the pandemic and inadequate follow-up of their mental health after discharge.
The results of this study concur with findings from a Canadian meta-analysis of 46 studies with a total of 47 677 participants who were pregnant during the pandemic (2). The prevalence of probable depression was estimated at 26 %, i.e. more than twice as high as pre-pandemic levels (10, 11). By way of comparison, the prevalence was measured at 11 % in a study based on data from the Norwegian Mother, Father and Child Cohort Study (MoBa) in the period 1999–2008 (7). In June 2020, a multinational study on perinatal mental health was conducted during the pandemic (12). The results indicated a prevalence of probable depression of 15 % among the Norwegian participants in the study, while the corresponding figure was 42 % for the United Kingdom. In June 2020, the pandemic had only been going for a few months, and most people in Norway thought it was over, while the United Kingdom had a significantly higher infection rate and was in lockdown. It cannot therefore be ruled out that the reference point may have had an impact on the relatively lower incidence in Norway than in the United Kingdom. In their study, depression was not measured in the same way as in our study, which makes it difficult to directly compare the two datasets.
Probable depression was defined in our study as a total score of 6 or more in EPDS-4. EPDS-4 is a screening tool that indicates depression, and is not a diagnostic tool. The Edinburgh Postnatal Depression Scale has a high level of sensitivity, but somewhat lower specificity, which may mean that false positives can occur, i.e. non-depressed women can be categorised as depressed. This means that the prevalence of depression is lower than the prevalence of women with a score above the threshold value (8). The scale is the most commonly used instrument internationally to screen for postnatal depression (13).
In studies based on voluntary participation, there is a risk of self-selection based on the degree of dissatisfaction (14). However, both the pandemic sample and the Ahus Birth Cohort sample were based on voluntary participation, which suggests that the risk of such self-selection is present in both samples. In the Ahus Birth Cohort, participants were recruited from a limited geographical area. Sociodemographic and reproductive characteristics nevertheless concur with data from the Norwegian Medical Birth Registry (15), which indicates that the Ahus Birth Cohort data are reasonably representative. The questionnaire in the Ahus Birth Cohort was distributed by post, while in the pandemic sample it was distributed nationally via social media and websites. Recruitment via social media does not necessarily lead to a selection bias. Previous studies suggest that recruitment via, for example, Facebook can provide representative samples on a par with traditional collection methods (16).
Several factors contradict the idea that the pandemic sample is more highly selected in terms of dissatisfaction and mental illness than the Ahus Birth Cohort sample. First, the proportion who were dissatisfied with their pregnancy experience was the same in both surveys. Second, about the same number of participants indicated pre-pregnancy mental health problems. There were more primiparous women in the pandemic sample than in the Ahus Birth Cohort, which may have had an impact on the results, as the risk of probable postnatal depression is somewhat higher among primiparous women (18). However, we found no significant difference in EPDS-4 scores with respect to parity. In the pandemic sample, a larger proportion of women had a college or university education than in the Ahus Birth Cohort. Since those with a higher education had a lower depression score, it is possible that the incidence of depressive symptoms in the pandemic sample is slightly underestimated. The disparities between education levels may be due to the fact that more people in this age group take higher education nowadays than was the case ten years ago (17).
Depressive symptoms were measured eight to ten weeks after childbirth in the Ahus Birth Cohort, while in the pandemic sample they could be measured up to thirteen months after childbirth. Some mothers in the pandemic sample thus completed the questionnaire just weeks after giving birth, while for others a significantly longer period of time would have passed between the birth and the reporting of current depressive symptoms. However, the risk of recall bias should be the same for all participants since they were only asked about depressive symptoms in 'the last seven days'. Nevertheless, there is a risk that the prevalence of depressive symptoms may be somewhat underestimated in the pandemic sample since most studies indicate that the risk of depression is highest during the first eight to ten weeks after childbirth, and then gradually decreases (18). Despite differences in recruitment and collection methods in the two samples, we believe it is important to report the findings.
Our findings reveal inadequate follow-up of postnatal mental health. However, it is not certain whether this is due to the pandemic alone or whether the pandemic exacerbated existing weaknesses in the mental health service provision and made follow-up even more challenging. A number of Norwegian studies suggest that the proportion of women with symptoms of postnatal depression may have been gradually increasing since the early 2000s (19). A new review article also suggests that follow-up of postnatal mental health was already inadequate prior to the pandemic (20). National Norwegian guidelines recommend that healthcare professionals ask pregnant women about their mental health (13). Our results reveal that almost a third were not asked about this.
The women in our study were significantly more dissatisfied with their stay in the maternity ward than the participants in the Ahus Birth Cohort study. It is conceivable that young women's expectations of care have changed over the course of ten years. Nevertheless, there is no doubt that the pandemic has impacted on hospital conditions, as the staff have had to spend time and resources on complying with infection control requirements. The strict visiting arrangements during the pandemic meant that partners were often sent home or had to wait outside until the birth was actually underway. Few partners have been able to be present in the maternity ward, which has meant a loss of important emotional support for the mothers. The staff have had to take over the partner's role, which may have led to extra stress for both the staff and the mothers. The proportion of infants receiving formula milk in maternity wards has increased during the pandemic (21), which may be due to the reduction in resources for helping mothers to initiate breastfeeding. One in three women in our study reported that their hospital stay was cut short due to the pandemic. The already short period in hospital (19) has thus become even shorter (22), and after discharge many women have spent their time in isolation.
Lack of follow-up during the ante-, peri- and post-natal periods is serious in terms of the woman's physical and mental health as well as from a broader societal perspective. It is a paradox that pregnant women receive close follow-up when it comes to physical aspects of their health, such as blood pressure, weight gain, glucosuria, etc., but are not systematically asked about their mental health despite it being well-documented that poor mental health can have just as serious consequences as poor physical health (23). The Norwegian government recently appointed a committee to propose specific measures to avert a mental health crisis in the wake of the pandemic (24). A total of 31 measures were outlined. Despite the fact that the pandemic has been very stressful for pregnant and postnatal women, it appears that the committee has forgotten about this vulnerable group.