Background: Parents of babies admitted to neonatal units (NNU) are exposed to a range of potentially distressing experiences, which can lead to mental health symptoms such as increased anxiety and post-traumatic stress (PTS). This review aimed to describe how anxiety and PTS are defined and assessed, and to estimate anxiety and PTS prevalence among parents of babies admitted to NNU. Method: Medline, Embase, PsychoINFO, Cumulative Index to Nursing and Allied Health literature were searched to identify studies published prior to April 14, 2021. Included studies were assessed using Hoy risk of bias tool. A random-effects model was used to estimate pooled prevalence with 95% CIs. Potential sources of variation were investigated using subgroup analyses and meta-regression. The review is registered with PROSPERO (CRD42020162935). Findings: Fifty six studies involving 6,036 parents met the review criteria; 21 studies assessed anxiety, 35 assessed PTS, and 8 assessed both. The pooled prevalence of anxiety was 41.9% (95%CI:30.9, 53.0) and the pooled prevalence of PTS was 39.9% (95%CI:30.8, 48.9) among parents up to one month after the birth. Anxiety prevalence decreased to 26.3% (95%CI:10.1, 42.5) and PTS prevalence to 24.5% (95%CI:17.4, 31.6) between one month and one year after birth. More than one year after birth PTS prevalence remained high 27.1% (95%CI:20.7, 33.6). Data on anxiety at this time point were limited. There was high heterogeneity between studies and some evidence from subgroup and meta-regression analyses that study characteristics contributed to the variation in prevalence estimates. Interpretation: The prevalence of anxiety and PTS was high among parents of babies admitted to NNU. The rates declined over time, although they remained higher than population prevalence estimates for women in the perinatal period. Implementing routine screening would enable early diagnosis and effective intervention. Funding: This research is funded by the National Institute for Health Research (NIHR) Policy Research Programme, conducted through the Policy Research Unit in Maternal and Neonatal Health and Care, PR-PRU-1217-21202. The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.
The review was prospectively registered with PROSPERO (CRD42020162935). NNU: This review includes parents of all babies admitted to NNUs for any level of care. Anxiety and PTS: Throughout this review, the term ‘anxiety’ is used to describe anxiety symptoms, which can vary from mild to severe17 or the presence of an anxiety disorder, such as generalised anxiety disorder. The term ‘PTS’ is used to describe PTS symptoms, which occur in response to an extremely negative or traumatic event, or the presence of acute stress disorder (from three days to one month after the event) or post-traumatic stress disorder (PTSD) (more than one month after the event). A search strategy was developed using a combination of free-text (title/abstract) keywords and MeSH (subject terms) to describe the key concepts of anxiety and PTS, parents, NNUs and prevalence. Medline, Embase, PsychoINFO, Cumulative Index to Nursing and Allied Health literature, Web of Science, ResearchGate and Google Scholar were searched (Appendix A). No date or language restrictions were applied. A search of grey literature was conducted using British Library EThOS, Open Grey and ProQuest Dissertations & Theses Global and studies. The websites of not-for-profit organisations Bliss and March of Dimes were also searched for relevant studies. In addition, the reference lists of all included studies were used to identify further relevant publications. The final search was conducted on April 14th 2021. Studies were included if they: 1) were cohort (prospective or retrospective) or cross-sectional in design; 2) assessed prevalence of anxiety and/or PTS at any time after birth; 3) included mothers, fathers, parents or other primary carers of babies admitted to a NNU. Two of three reviewers (RM, SH, FA) independently screened the titles and abstracts of all studies identified by the search. Full texts were independently screened by two of four reviewers applying the review eligibility criteria (RM, SH, HB, FA). Screening was performed using Covidence software.18 Disagreements regarding study eligibility were resolved through discussion and consensus within the review team. Study authors were contacted if cut-off points were not reported, if anxiety and/or PTS data were reported as mean scores or combined with prevalence of depression, or if the study was available as an abstract only. Additional information was provided by 14 study authors. Where no data were obtained from authors, missing data were recorded as not reported and subsequently excluded from meta-analyses and meta-regressions. The following data were extracted for each included study: year and country of publication, study objective, study design, study period, NNU level, infant length of NNU stay, study inclusion/exclusion criteria, demographics of parents and babies, assessment tool, cut-off point, time of assessment and prevalence. Risk of bias was assessed using the Hoy risk of bias tool (Appendix B).19 The tool consists of ten items: items one to four assess external validity (selection bias (items 1-3) and non-response bias (item 4), items five to ten assess internal validity (measurement bias (items 5-9)) and analysis bias (item 10). Appraisal of each item provides a subjective assessment of risk of bias as low, high or unclear. All data were extracted and independently cross-checked by at least two authors (RM, SH, HB, FA). Pooled prevalence estimates of anxiety and PTS were calculated by combining estimates from each study. Meta-analysis was conducted using the “Metaprop” function in STATA 15.9.20 A random-effects model was applied and the results were reported as proportions with 95% confidence intervals (CI). The data were analysed by time of assessment: up to one month after birth; from one month to one year after birth; and more than one year after birth. If studies reported more than one assessment of anxiety/PTS in the time period, the time point with most participants was included. Subgroup analyses were pre-specified and conducted across the following study characteristics: setting (high vs. middle income countries), design (cohort vs. cross-sectional), sample representativeness (low vs. high risk of selection bias on item one of the quality assessment vs. no on item one), anxiety measurement tool (self-report vs. clinical interview), sex of parents (male vs. female), self-report scale (State Trait Anxiety Inventory (STAI) state scale vs. others for anxiety; Perinatal Post-Traumatic Stress Disorder Questionnaire (PPQ) vs. others for PTS) and prematurity level (< 33 vs. ≥ 33 weeks gestation). Planned subgroup analyses by birthweight, neonates that had surgical procedures, level of NNU, and length of stay were not feasible due to insufficient data. Sensitivity analysis based on study quality was also planned but could not be performed because no study was low risk of bias on all items. Evidence of variation in anxiety/PTS prevalence due to between-study heterogeneity was assessed using the I2 statistic, which describes the percentage of variation not due to sampling error. An I2 value above 50% indicates moderate heterogeneity and above 75% indicates high heterogeneity.21 Where there was evidence of high heterogeneity and there were sufficient numbers of studies, meta-regression was performed to investigate whether any variation in prevalence estimates was explained by study characteristics. Patient and public involvement (PPI): A voluntary group of parents, whose babies received care in a neonatal unit and a representative from Bliss provided the PPI input for this review. Parents provided a personal insight from their experience of having a baby cared for in a NNU and worked collaboratively with the team to improve the readability of the scientific language used in the review and interpreting the findings to a wider dissemination of the findings. The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.