Semi-structured and focused group interviews

Twenty-eight ‘key stakeholder’ participants took part in the study. Semi-structured individual key informant qualitative interviews were undertaken with 9 senior leaders/decision: each responsible for one or more facets of the major management decisions during the outbreak. Additional data were collected with 19 frontline HCWs (10 nurses and 9 physicians) who were each in direct patient care contact with MERS-CoV patients during the outbreak. Recruitment was undertaken via volunteer quota sampling of decision-makers and staff from across the hospital. Data were collected in the clinical institution by independent researchers (AC) and (HK), with no direct involvement with the site, with schedules developed from past literature, approaches to learning organisations, and the respective role(s) of participants (online supplementary appendix 1). Interviews were audio-recorded via a digital device with data transcribed immediately after data collection. Each interview/group lasted between 45 and 100 min; transcripts included non-verbal behaviour. Prior to the interviews, institute revenant written documentation such as memorandums and committee meeting minutes was analysed to direct the questions posed during the qualitative data collection (table 1, online supplementary appendix 1).

Data management

Qualitative data using process evaluation generates insights from key stakeholders on contributions to, and factors affecting key outcomes or processes, in organisations or interventions.14 15 Interview transcripts were analysed manually to determine common themes using recognised principles of qualitative research.16 For each interview, codes, subthemes and themes were then identified and subsequently refined to avoid redundancy and guarantee accuracy via a cyclical analytical process by two of the investigators.16 The analysis moved back and forth between the interviews to ensure the finalised analysis was completed.

Rigour was maintained via a variety of recognised techniques.16 Member checking of the qualitative data enhanced credibility and transferability: the results were presented to 13 participants to ensure the findings were comprehensible and had resonance. Participants concurred with the proposed results. Peer debriefing was used in which a second researcher analysed a random selection of interviews to ensure themes were understandable. No major changes to the analysis arose from this debriefing.

Patient and public involvement

The research question and design were informed by the high mortality and severe morbidity resulting of MERS-CoV. Neither patients nor public were involved in the study; answering the research question does not require their participation.

The high stress of the outbreak

All participants reported that the MERS-CoV outbreak was seen to constitute an episode of the most pressing and serious clinical significance to the organisation and country. Consistently across interviews, it was described in such terms as a serious and straining situation and a situation that carries national and international significance .

Frontline HCWs and senior decision-makers alike not only perceived this high gravity, but also experienced extremely high and diverse demands personally ( a very demanding situation) . Throughout the interviews, participants described the outbreak as being a sustained period of severe and sustained ‘tension’, ‘doubts’, ‘challenges’, ‘fear’ and ‘anger’. Frequent visits, communications and collaborations from external regulatory agencies (such as the Saudi Ministry of Health, the US Center for Disease Control and Prevention and the WHO), were perceived to be useful but also compounded these stresses.

The perceived seriousness of the outbreak was reported to motivate staff to expend maximum effort to assist in its management. However, the high levels of stress were exacerbated by a pervading sense that controlling the outbreak involved hard and uncertain progress . In the midst of the outbreak, both how it would progress and the future were seen to be very unclear.

Factors perceived to have contributed to wide occurrence of the outbreak

Strong consensus existed among participants that the organisation’s growing reliance on emergency department boarding was the main contributor to the outbreak. While some participants lamented that this build up was ‘permitted’ by senior decision-makers to occur over the long-term, other participants saw other upstream factors as also being influential, notably inadequate implementation of Infection Prevention and Control (IPC) guidelines. Indeed, poor recognition of the importance of IPC principles across the organisation was perceived to be compounded by the high demands placed on the ED by the large patient population. Other more upstream factors seen to contribute to the outbreak included: high trust patients had for the National Guard Health Affairs, the relative lack of primary care services, problems associated with patient flow across the city, poor communication and teamwork between the city hospitals, and a lack of pre-emptive national planning to allow hospitals across different sectors to share the burden of the growing city population.

Factors perceived to have contributed to the success of outbreak control

Throughout this challenging situation, teamwork and collaborative management were seen to be pivotal. There was a sense of close alignment between both decision-makers and frontline HCWs in management priorities. The approach of senior management to the situation was perceived by participants as being open, without blame or ‘finger-pointing’. Indeed, managers were seen to have focused on empowering HCWs to work collaboratively to address the outbreak, involving staff in outbreak control and stimulating staff resilience and teamwork. In this way, the dominant culture was perceived by staff to be one of collective leadership practice (table 1) with leaders being seen to have practiced high levels of availability, visibility, empowerment of middle management and strong links with frontline staff during the height of the outbreak.

Key factors seen to contribute to this sense of collaborative culture included: the existence of a clear shared-vision across staff of the high priority outbreak control measures, the high frequency of meetings of decision-makers involved in the outbreak control (twice daily), rapid and efficient decision-making, involvement of all the right units and decision-makers in decisions, and a high level of accountability. Having a centralised Command and Control Centre committee was seen widely to facilitate openness with high accountability, strategic utilisation of team diversities and strong mutual support. The success in controlling the outbreak in a relatively short time fostered a widespread sense of pride among those working in the organisation. Collectively, these measures served to improve mutual trust between front liners and top management .

Factors inhibiting outbreak control

Factors perceived to inhibit management of the outbreak were mostly related to initial poor management practices and the negative and compounding effects of media reporting and high stress. The organisation was perceived to be acutely and chronically slow in responding to the outbreak. For example, organisational responses from management to the outbreak was perceived to be relatively slow compared with the rapid speed with which the infection spread. Moreover, despite references to high management transparency, some staff cited that poor staff orientation and management ambiguity contributed to staff being isolated and unclear about the decisions and measures being adopted by senior management to promote IC. This perceived poor communication created additional confusion that led to low coordination of IC instructions from bothIC and the nursing teams. Consequently, some health professionals, and other administrative and military staff, did not even appreciate the severity of the outbreak. Their consequent lack of attention to reducing the number of new patients and visitors entering the institution was seen to further elevate infection risk.

All participants referred to consistent and pervasive negative media commentary on the MERS-CoV outbreak occurring at the KAMC-R. This negative coverage contributed to significantly negative public perceptions of the Ministry, KAMC-R, its senior decision-makers and the frontline HCWs. These negative commentaries were evident across local mass media (television, radio and newspapers) and social media - particularly Twitter. The negative media reporting was cited as negatively impacting staff morale and affecting workers socially, psychologically, and mentally. During this challenging period, this compounded the negative effects of work demands. Reactive and poor media management by the KAMC-R was seen to contribute to the ongoing frequency and negativity of this commentary. Participants suggested that the institution should have a media centre to coordinate media coverage in such emergency circumstances (table 2).

Staff capacity to handle this challenging situation was further reduced given the perceived wide prevalence of high anxiety in staff due to the lack of appropriate staff counselling and mental health support . The psychological demands on frontline health professionals escalated due to a range of coalescing factors; most notably the need to manage the outbreak simultaneously with the closure of most of the institution’s units, combined with the negative media commentary. Almost all frontline participants strongly and repeatedly expressed the need for counselling and mental health support of employees.

Research limitations and strengths

This research documented learning from an unusual, burdensome and serious infection outbreak and used methods recognised to be well suited to explore the complexities of outbreaks.31 32 Similar to other ‘rapid’ qualitative accounts of major outbreaks with high mortality notably ebola,31 this study addresses key interdisciplinary aspects of perceived outbreak causes, infrastructure, IC, facilities and health needs.32 Unlike this previous work, the rigour of this study was increased via comprehensive details of the sample, participants and context.32 As with other studies,32 participants included in the study were very well-placed to provide insights into the outbreak but inevitably then the research was also conducted retrospectively, was based on subjective data, and involved a select group of participants whose perspectives may differ from the broader population. As with other studies,32 while independent researchers undertook the interviews, responses may have been influenced by perceived ‘official’ links between the project and the organisation. Staff perspectives may have also been influenced by the time duration since the outbreak was curtailed (around 8 months). We addressed these limitations through different data collection sources, member checking and data triangulation.