Coronavirus (COVID-19): care home outbreaks - root cause analysis

4. Early warning systems

The current RAG reporting system from the DPHs did not identify process issues deteriorating before the outbreak occurred in each of the four homes in this review. The RCA indicated that identifying an outbreak was dependent upon the care home manager identifying an outbreak as a possibility and acting quickly, or in response to timely reporting of testing data from the HPTs to the care home. In each circumstance there was a delay in the first case being identified and some days between that and an outbreak being declared, in the most part this was a result of testing turnaround times for confirmation of two or more cases as an outbreak. Once an outbreak was declared control measures were reported as being put in place, however at this point there had been likely widespread transmission.

There was a view expressed that given the DPH report is based on TURAS safety huddle data, which is self-reported by care homes, it can only be as good as the interpretation by the reporter and the content submitted by that individual. Again a process dependant on an individual is not a system and raises issues of resilience. Phone call follow up alone was also identified by the DPHs as insufficient; visiting and observing is key to understanding that effective control measures are in place at care home level.  There is a need for early warning indicators to be sensitive to those factors which have impact for IPC. These include: staffing (% of agency use and movement of staff from other homes or the NHS, in addition to ratios of registered and carer staff to residents), sickness absence, increased testing initiated by the care home, PPE and test kit availability or the need to order extra supplies. The RCA also identified the potential of community prevalence (inclusive of neighbouring boards where the staff live) as an early warning indicator. 

As already mentioned, recognition by care home staff, of the wider spectrum of COVID-19 symptoms in older people, to ensure there is a high degree of suspicion is also key. One home reported that a general deterioration of residents, such that they required hospitalisation was the first noticeable sign, and as there was no cough, fever or loss of taste and smell reported by those residents, the staff did not suspect COVID-19 as the underlying condition. One of these residents was tested on admission to hospital, however the result was not reported back to the care home, which negated an opportunity to intervene with control measures earlier.

NHS boards reported that in this second wave they are identifying outbreaks earlier and managing more quickly and responsively, however in these outbreaks there was a delay between first known or suspected case, confirmation of cases and declaring an outbreak noted in the safety huddle data. The delays in each of the circumstances were a result of a variety of issues in different parts of the system: delayed recognition of cases because of low index of suspicion (not familiar with broader syndrome of COVID-19 in older people), delayed identification of cases related to limited testing availability at the right time, asymptomatic/pre-symptomatic residents and staff members, delays to reporting testing results to care homes. This may mean that there were particular conditions, which made these homes different to other care homes, and as a result these homes may not be representative of the wider, current picture across Scotland. 

A review of the TURAS safety huddle system data indicated that there were potential early warning indicators in each of these four homes, which could be an opportunity to intervene early. These included sickness absence of staff and indicators of staff shortages, supplementary staffing, falls, deteriorating residents, IPC indicators. The delay between the first positive case and confirmation of the outbreak noted in the data could also have been a trigger for action. As already mentioned, it is important to note that these data are self-reported by the care home, and whilst these have not been subject to any quality assurance or validation, there is value in pursuing this tool and assessing its merits as an early warning indicator in the care home system. To enable this, these data need to be reviewed internally within the care home on a daily basis by the care home staff themselves, with swift action taken to prevent outbreaks where possible, and to minimise the number of cases involved in outbreaks otherwise. Care home staff may need support and motivation to undertake such action and it would be important that this is done in partnership with the care home oversight group and HSCP. We noted that TURAS returns in relation to staffing elements are not sensitive enough to provide detailed information on staffing levels; whilst overall staffing numbers are reported, they are not defined in terms of shift length and so from the information submitted, assurance in relation to actual staff on duty at any one point in time is not able to be ascertained. Delay, or inability to recognise infection control deficiencies (which in some homes had been self-assessed as 100% compliant before and during the outbreak, despite challenges with staffing), and control measures being put in place after more than two cases were confirmed, were common to all of these outbreaks. Initiation of the escalation process in response to these early indicators or after a single confirmed case may have prevented the high attack rate in these homes. 

Recommendations:

• TURAS, and supporting processes for its use in the HSCP and care homes, should continue to be further developed to ensure it can be used as effectively as possible as an early warning system 
• Care homes should be supported to use the TURAS data for local improvement
• A further detailed review of staffing rosters and workforce capacity should be considered based on the findings from the TURAS indicator data, it may be helpful for care home oversight groups to work collectively with care homes in the use of workforce tools to enable system level planning and mutual support