We followed the SPIDER (Sample, Phenomenon of Interest, Design, Evaluation, and Research Type) framework used in the original review (14).

Our sample included maternity HCPs directly involved in provision of maternity care during the COVID-19 pandemic. A range of professions were captured, including, but not limited to midwifery, nursing, and obstetrics. Whilst we sought studies published globally, in this review, we have restricted our sample to UK-based studies (see Search Strategy and Selection section for further detail). The phenomenon of interest was HCP experience of maternity care provision during the pandemic, including all antenatal care (except abortion), labour and childbirth, and up to six months postpartum. Care in all settings was considered. Qualitative study designs of interest included: descriptive, exploratory, and interpretive studies; ethnographic studies; observational or mixed-methods studies in which qualitative data had been extracted separately; survey designs with open-text questions when significant qualitative data had been collected and formally analysed; linguistic studies; and studies of public discourse. Only published literature was included.

Literature published between 01 June 2021 and 30 September 2023 [building on the previous review's (14) search, 01 January 2020 to 13 June 2021] was sought. The search strategy was restricted to English language.

Systematic searches were undertaken of the electronic databases of Scopus, MEDLINE (Online counterpart of MEDLARS MEDical Literature Analysis and Retrieval System), EMBASE (Excerpta Medica dataBASE), CINAHL (Cumulative Index of Nursing and Allied Health Literature), PsycINFO and the Cochrane COVID Study Register. The search terms and keywords used in the Flaherty et al. review (14) were adopted (Supplementary Table S2).

EndNote Reference Manager was used to clean search result duplicates, and citations were uploaded to the Rayyan web-based systematic reviewing tool. Team members (TD, LP, GH, MW, SAS, HDM, PvD, LAM) independently screened each title and abstract, followed by full-text review. After each screening stage, disagreements were resolved through discussion with the wider team. Given the large number of studies meeting inclusion criteria, and the focus of RESILIENT on maternity care in the UK, a decision was made prior to data extraction to sub-divide the review by population of interest (women and birthing people or HCPs) and geography [UK, other high-income countries (HICs), or low- and middle-income countries]. The remaining studies have (19) or will be synthesised separately.

Data were extracted independently by two reviewers (EB, YP) into a pre-designed Microsoft Excel data extraction sheet, and checked by wider-team members during regular discussions. Extracted information included: characteristics of studies (e.g., reference, aims, setting, and dates of data collection) and participants (e.g., number, setting), data collection method, details of analyses, and themes identified, all taken from the results sections of included papers. Then, each paper was imported into NVivo qualitative research software for coding and synthesis of Discussion sections. Of note, Results sections were not coded to avoid replicating codes/themes and rendering logic circular.

In line with the previous reviews (7, 14), methodological quality was assessed independently by two team members (EB, YP) and checked for correctness by other authors, using an adapted version of a 12-item EPPI-Centre (Evidence for Policy and Practice Information and Co-ordinating Centre) tool which captures information on the reliability and validity of study methods and reporting, for qualitative evidence synthesis (20). Data were included for synthesis, regardless of quality, to provide relevant ‘views/experiences’ data. We did, however, interrogate our final results to ensure that inclusion of low-quality papers, if any, did not compromise the integrity of resulting themes (that is, the removal of codes derived from low quality studies was not found to affect the overall set of derived themes).

Thematic Synthesis (21) was undertaken based on a set of a priori concepts which address RESILIENT aims: (1) Care-seeking and care experience, (2) Virtual care, (3) Self-monitoring, (4) COVID-19 vaccination, and (5) Ethical future of maternity care services. Extracted data from each study were aligned with one or more of these key concepts, then data under each concept were coded, and descriptive themes generated inductively. Data were synthesised independently by two reviewers (EB, YP), to ensure cohesion and congruity in coding, with regular discussion to resolve any conflicts and agree on derived themes.

Figure 1 illustrates the search and selection process (18). The initial literature search yielded 21,860 records. Records were removed if they were: duplicates (n = 2,925); ineligible at title/abstract screening (n = 18,468); could not be retrieved (n = 54); or were ineligible at full-text review (n = 200; see Figure 1 for reasons). Thereafter, 215 records met inclusion criteria, of which 15 studies of UK HCPs’ experiences of delivering maternity care during the pandemic are reported here (4, 5, 22–34).

The 15 included studies were exclusively UK-based, apart from one; this compared care between the UK and the Netherlands (31), and presented data all together. One study (18) presented data from maternity and children's healthcare professionals, commenting on each set of professionals.

There were 940 participants, with additional unspecified numbers from 224 maternity units (23, 30) and eight maternity care policy organisations (30). Data collection ranged from February 2020 to November 2021, extended by 11 months beyond the original review (14). Study methodology varied: semi-structured interviews (n = 8) (5, 22, 24–26, 28, 29, 34), semi-structured interviews alongside analysis of policy and guidance documents (n = 2) (30, 31), mixed-method survey with free-text analysis (n = 3) (23, 27, 32), both interviews and a mixed-method survey (n = 1) (4), and a focus group with midwives (n = 1) (33). For data analysis, most studies utilised thematic analysis (n = 7) (22, 25–28, 32, 33); other methodologies included: content analyses (n = 1) (30), descriptive analysis (n = 2), 23,34 framework analyses (n = 2) (4, 31), or grounded theory analyses (n = 3) (5, 24, 29). For detailed characteristics of included studies and their key findings, see Supplementary Table S3.

Study quality varied. Eight studies met all 12 quality criteria (5, 22–24, 29, 31, 33, 34); two studies met 11/12 criteria (27, 30) as they did not actively involve participants in the design and conduct of the study; one study met 10/12 criteria (4), two studies met 9/12 criteria (26, 28), one study met 8/12 criteria (25), and one study met 7/12 criteria (32). As such, all studies were deemed to be of moderate-high quality; for details, see Supplementary Table S4.

Our synthesis identified nine themes within three of the five RESILIENT concepts: care-seeking and care experience, virtual care, and ethical future of maternity care services (Supplementary Table S5). No studies reported on the RESILIENT concepts of self-monitoring or COVID-19 vaccination. Passages of text from the original discussion sections are presented in Tables 1–3 to support synthesised findings.

Fourteen studies (4, 5, 22–34) contributed data to this concept, with four themes: 1.1 Changes to existing provision of care, 1.2 Limitations placed on the partner, 1.3 Mental health and lack of support networks, and 1.4 Barriers to implementation of reconfiguration strategies (Table 1 for supportive quotations).

Nine studies (4, 5, 23, 25, 27, 28, 32–34) contributed data to this concept, describing how some maternity care changed from in-person to virtual, by telephone or video-conference. Three themes were identified: 2.1 Impact on quality of care, 2.2 Increased convenience and flexibility, and 2.3 Digital exclusion (Table 2 for supportive quotations).

All fifteen studies (4, 5, 22–34) provided data for this concept, describing how maternity services should be built back in a fairer and ethical way, to prevent further exacerbation of health inequities. This concept was coded into two themes: 3.1 Optimising patient care, and 3.2 Service users and staff as the driving force for change (Table 3 for supportive quotations).

This systematic review of 15 qualitative studies (4, 5, 22–34) of HCPs’ experiences of providing maternity care during the COVID-19 pandemic in the UK, builds on a previous qualitative evidence synthesis (14).

Key findings included reduction in provision of community midwifery services leading to perceived loss of autonomy for women and birthing people, challenges with providing good quality care, and inadequate guidance and support for staff regarding protocol changes and safety measures. Altered care-provision and limitations placed on the role of fathers, partners, and non-gestational parents during appointments and birth were perceived by HCPs to be detrimental for women's emotional and physical wellbeing. Staff reported loss of morale, unrealistic work pressures, and reduced staffing—making it difficult to successfully implement reconfiguration strategies.

Studies evaluating experiences of virtual care highlighted HCPs’ concerns about care quality, compared to in-person care, especially for high-risk and vulnerable groups. Participants felt that comprehensive mental health and wellbeing assessments cannot be completed virtually, potentially jeopardising women and birthing peoples’ safety, and leaving staff feeling they were unable to fulfil their duty of care. Access to digital devices and reliable internet connectivity were highlighted as problematic. However, HCPs expressed that virtual care increased convenience and flexibility, and some HCPs found it easier to provide continuity of care and more frequent contact with women and birthing parents.

HCPs perceive an ethical future for maternity care services in the UK to include: personalisation of care to suit individuals’ needs, the offer of in-person care when necessary, and the offer of hybrid care for others who prefer to avoid coming to hospital. The synthesis emphasised the importance of a co-designed and collaborative approach to designing future maternity care, by including both service-users and HCPs in the decision-making process.

To our knowledge, this is the only UK-focused systematic review of HCPs’ qualitative experiences of delivering maternity care during all three COVID-19 pandemic lockdowns.

We add 15 UK publications (4, 5, 22–34) to the single UK study included in the previous qualitative thematic synthesis by Flaherty et al. (14), comprehensively enhancing our understanding of the impact of the COVID-19 pandemic on UK HCPs’ experiences of providing care during an unprecedented health system shock. Adhering to the aims of RESILIENT, data were synthesised according to our five pre-defined concepts; however, our findings resonate with those of HCPs internationally. The six themes identified in Flaherty et al.'s review drawing on the global literature align primarily with our core concepts of Care-seeking and care experience: altered maternity care, altered care structures and provision, capacity to provide care, professional and personal impact, professional impact, and personal burden (14). We expand, by adding themes related to virtual care and, importantly, HCPs’ views on an ethical future for UK maternity care services (7). However, no data were found to align with the RESILIENT concepts of Self-monitoring or Vaccination.

Our findings align with those of service-users and specific groups of HCPs studied by other researchers. An online survey of parents in Northern England found a reduction in women and birthing people's choices and autonomy over their care (particularly with respect to birth-planning), which jeopardised their overall satisfaction and wellbeing (35). Others reported how it was difficult for HCPs to work in ways which incorporated infection control measures, whilst meeting the needs of women and birthing people, particularly given restricted personal engagement and the ability to provide supportive touch (36, 37). The negative consequences of restrictions placed on fathers', partners', and non-gestational parents’ involvement have been echoed in several other works (6, 38, 39), including potentially reducing these individuals’ ability to bond with their baby, and to offer support to the mother or gestational parent (40)

A key finding of our review was the challenge faced by staff in fulfilling their duty of care, in the face of staff shortages and limited resources. This issue has been recognised and debated by the UK Government (41). Staff surveys in the global setting have attributed staff shortages to heightened stress levels and burn-out (42). Several studies document an increase in depression, anxiety, and stress among HCPs during the pandemic, along with post-traumatic stress symptoms (43, 44).

Our findings that staff had concerns about developing trusting and meaningful relationships with women and birthing people through telephone or video consultation was echoed by service-users, who felt virtual antenatal consultations provided impersonal care and had a negative impact on how much information women and birthing people chose to disclose to their HCP (45). Workforce surveys and those from the UK's communications regulator have associated digital poverty during the COVID-19 pandemic with disabilities and lower socioeconomic background and housing tenure (46). An extensive narrative review of telemedicine in the United States of America emphasised the need for equitable access to digital technology, as well as its potential (47). With a global shift to virtual care delivery in a post-pandemic world, it is crucial to carefully consider the ramifications of using digital technology for groups that are already marginalised and prone to digital exclusion (34). It is crucial to understand the multilayered aspects involved in the adoption and implementation of this technology, from the perspective of all stakeholders, and it should not replace traditional face-to-face care, but rather complement it (51, 52). Finally, the strong desire of women and birthing people to have a model of maternity care that supports women-led decision making (48) speaks to the collaborative working and co-design expressed by HCPs in the literature we reviewed.

Although the roll-out of the COVID-19 vaccination program began in the UK in December 2020, and management of routine self-monitoring of symptoms for pregnancy complications (such as gestational diabetes and hypertension) were major maternity service reconfigurations during the pandemic, the paucity of literature reporting HCPs’ views on these concepts may indicate that staff did not perceive them to have had a major impact on their day-to-day lives, or that this was not a research priority in studies with HCPs. Other research with service users has shown self-monitoring of symptoms during pregnancy to be associated with implementation and resourcing issues (49). Within the RESILIENT programme of work, we have found in interviews with HCPs, policymakers, women, and partners that vaccination, particularly mandatory vaccination programs for staff was a contentious issue (50). We would endorse future research to confirm these findings within other contexts.

This review benefits from robust data extraction and synthesis, with all studies screened for inclusion independently by at least two authors; while data extraction, quality-assessment, and synthesis were conducted independently by two authors. We evaluated HCPs’ experiences of care provision over all three pandemic lockdowns in the UK. Importantly, we gathered information on how HCPs believed maternity care can be improved in the future, and their emphasis that these should be informed by their own experiences. Given the sheer volume of literature published in the last three years about the impact of the pandemic on experiences of maternity care, and the focus of RESILIENT on the pandemic in the UK, we limited the scope of this review to UK studies only. Nevertheless, Flaherty et al. (14) reported similar views of HCPs in their review of global literature; with no thematic differences between HICs and LMICs, and we plan to complete the additional RESILIENT study systematic reviews in describing longer-term experiences internationally, imminently. Whilst we can take these findings from the UK as a critical case (53) from which we can extrapolate to other settings we realise they may not be wholly generalisable to HCPs’ experiences in other parts of the world, particularly where the system is not modelled on being ‘free-at-point-of-use’. Future publications of the RESILIENT study as well as other research should focus on comparing experiential literature between different healthcare settings. Finally, although our search strategy for the population of interest included a range of professional roles within maternity care, it may not represent the whole maternity workforce in the UK, particularly as specific social determinants (such as gender, ethnicity, or geographic location of individual Trusts) were not considered. This work would be complemented by further local and context-specific research.