Multi-disciplinary community respiratory team management of patients with chronic respiratory illness during the COVID-19 pandemic

Patient demographics

About 516 patients were referred to the CRRT in May 2020. All patients were residents within the NHS GGC health board area. COPD was the commonest primary diagnosis comprising 65% of patients referred, and this increased to 72% when COPD/asthma overlap patients were included. The second most common diagnosis was asthma (12%), followed by interstitial lung disease (ILD) (7%), with smaller numbers of patients with other respiratory diagnoses (Table 1). Diagnoses classed as “other” included obstructive sleep apnoea, pulmonary embolism, and lung cancer. The mean age of patients referred was 69. 340 (66%) patients were female and 176 (34%) were male. Females outnumbered males in all primary respiratory diagnoses with the exception of ILD, in which 21 of the 36 patients were male (Table 1).

Table 1 Patient demographics n = number and overlap refers to a primary respiratory condition with features of both asthma and COPD.ED attendance, hospital admissions, and death rates

Sixty-six patients (13%) attended ED within 28 days of referral to the CRRT. Fifty-five patients (83%) who attended ED were subsequently admitted to the hospital, suggesting a genuine need for inpatient care among the majority of patients who attended ED. Eight patients were admitted directly to the hospital without attending ED, meaning a total of 63 patients (12%) were admitted to the hospital (Table 2 and Supplementary Table 1).

Table 2 Patient Demographics and Outcomes by CRRT Triage Pathway % relates to proportion of pathway.

Twenty-five patients (4%) died within 28 days of referral. The mean number of days from referral to death was 11 (range 0–28, median 12, IQR 3–15). The mean age at death was 78 years (range 53–98, median 79, IQR 72–84). Fifteen patients who died were female and ten were male. Eight (32%) patients who died within 28 days were nursing home residents. Deaths occurred in 11 patients with COPD (3% of COPD patients), one patient with COPD/asthma overlap (3% of overlap patients), one patient with bronchiectasis (10% of the group), and five patients identified as end-of-life (56%). Two patients died in the other/unknown group (12%), both of whom had metastatic malignancies. Six patients died after a COVID-19 diagnosis, three of whom were triaged with COVID-19 as the primary diagnosis, and five patients had confirmatory virology testing. No deaths occurred in patients with a primary respiratory diagnosis of asthma.

Utility of the triage pathway

19% of patients were triaged as red, 46% as amber, 27% as green and 8% as blue (Fig. 1). Scoring the red, amber and green pathways according to triage severity (where red is highest and green is the lowest severity), attendance, admission and death correlated negatively (R = −0.98, −0.96 and −0.97 respectively), i.e. the higher the triage severity, the smaller the proportion attending or being admitted.

The red pathway had the highest proportion attending ED (21%) and this was significantly higher than the amber and green pathways (p = 0.03 and p = 0.004, respectively), but not the blue pathway (p = 0.419). As only seven patients were admitted directly to the hospital without attending ED, the proportions of the 63 patients admitted were similar—21 (21%) red, 26 (11%) amber, 11 (8%) green and five (13%) blue pathway patients (Table 2). The highest number of deaths were in the blue “end-of-life” triage pathway (P < 0.001 when compared with all other pathways), followed by the red “high risk of deterioration” pathway (Table 2). The one patient who passed away in the green triage pathway died from acute abdominal sepsis, meaning no patients triaged as green subsequently died from a respiratory cause.

Utilisation of remote consultations

The total number of CRRT consultations in the study population was 2261. Of these, 1971 (87%) were conducted remotely via telephone or an online video calling platform. For their initial consultation, 431 (84%) patients were assessed by remote consultation and 85 (16%) at a home visit. About 181 patients (35%) had at least one home visit during follow-up (mean 1.6; range 1–8; median 1.6, IQR 1–2), meaning 335 patients referred (65%) were managed entirely remotely.

Patients received, on average, 4.4 consultations (range 1–44; median 3; IQR 1–5). A higher number of consultations was not significantly associated with ED attendance (Pearson’s correlation r = 0.036, p = 0.416), hospital admission (r = 0.074, p = 0.095) or death (r = −0.085, 0.055). This was also true when the blue pathway data were removed from the analysis. 33% of patients who presented to ED had only one CRRT consultation, while just 22 (13%) of the 169 patients who had only one consultation attended ED.

The proportion of patients receiving a home visit correlated with triage category severity. A significantly greater proportion received a home visit in the red pathway; 71% of patients compared with 32% of amber patients (p < 0.001), 18% of green patients (p < 0.001) and 28% of blue patients (p < 0.001). Red pathway patients were also more likely to receive a home visit as their first consultation, with 30% of red pathway patients receiving a face-to-face review at home, compared with 12% of amber (p < 0.001), 15% of green (p = 0.005) and 13% of blue patients (p = 0.037). Of note, a large number of blue pathway patient referrals were requests for palliative oxygen delivery rather than requests for patient review; this includes ten (71%) of the 14 blue pathway patients who died and 38% of all patients who died within 28 days of referral.

Patients seen face-to-face initially after referral were not significantly more likely to attend ED (15% of initial home visits vs 12% of remote consultations, p = 0.388), be admitted to hospital (15 vs 11%, p = 0.288), or die within 28 days (5 vs 4%, p = 0.802).

Importantly, no nosocomial COVID-19 infections were identified due to CRRT input among CRRT staff or patients managed at home.

Staffing and cost analysis

In total, the CRRT employed a total of 26 nurses, 12 physiotherapists, one occupational therapist and two respiratory consultants. Of the staff employed, 25 worked full time for the CRRT while the others either worked part-time or combined work for the CRRT with other duties. 93% of referrals were made on weekdays, with an average of 24 referrals per weekday and 4 per weekend day. Staffing was adjusted accordingly, with an average of 23 staff working each weekday and 13 each weekend to cover new referrals and follow-up consultations. The team provided an average of 490 consultations per week, meaning each staff member averaged 14 consultations per day, including telephone consultations and home visits. Respiratory consultant input to the daily virtual ward rounds required approximately three sessions per week with an annual cost estimated at £28,000 per annum.

Based on the above staffing levels, the cost of staffing the service across the NHS GGC was approximately £48,789 per week, or £2,543,815 per annum (Table 3). Therefore, the average cost was £86 per CRRT consultation and £378 per patient referred in May 2020. This is compared to an average cost of £3602 per patient for a secondary care admission in NHS GGC in 2018–1919 and £3000 per COPD-related inpatient stay in NHS Scotland20.

Table 3 Projected costs of staffing the CRRT based on staffing levels required during the month of May 2020.Effect on COPD ED attendances

Patients with COPD were the most represented group referred to the CRRT (Table 1 and Supplementary Table 1). Prior to implementation of the CRRT, increases in COPD EAs over time were evident in GGC, while the trend for all-cause EAs was approximately flat (Fig. 2). Conversely, in RoS, COPD EAs were decreasing, and all-cause EAs saw a slight rise over time. Following the start of the pandemic and the establishment of the CRRT service, a large drop was seen in both GGC and RoS in all-cause EAs and COPD EAs. This was then followed by a sharp increase, reduction and second increase in all-cause EAs (likely reflecting the waves of the pandemic as people attended ED less during lockdown periods). COPD EAs did not, however, see as large an increase in either area (Fig. 2). Adjusting for changes in COPD EAs in RoS, there is a significant decrease in the trend in COPD EAs in GGC (RR = 0.96 (0.94, 0.98) for each additional month under the Poisson model) compared with the counterfactual, i.e. if the service had not been in place (Fig. 3 and Table 4).

Fig. 2: COPD and all-cause emergency attendances (EAs) per month, January 2018–May 2021, in Greater Glasgow and Clyde (GGC) and the Rest of Scotland, excluding Fife, Lothian and Tayside (RoS).figure 2

a COPD EAs per month, for residents of GGC and RoS. b All-cause EAs per month for residents of GGC and RoS. The shaded area represents the phase-in period of the community respiratory service.

Fig. 3: The effects of the CRRT on COPD emergency admissions.figure 3

Segmented linear regressions of COPD EAs in Greater Glasgow and Clyde with the Rest of Scotland, excluding Fife, Lothian and Tayside (RoS) as control and adjusting for all other cause emergency attendances (EAs), January 2018–May 2021; deseasonalized linear trend under model and the predicted trend had the community respiratory project not been implemented.

Table 4 Parameter estimates, standard error, relative risk and 95% confidence intervals, and p values from the segmented Poisson regression predicting COPD EAs, adjusting for seasonality.

留言 (0)

沒有登入
gif