The Alfred Health post‐COVID‐19 service, Melbourne, 2020–2022: an observational cohort study

Some people experience symptoms that persist for weeks or months after acute coronavirus disease 2019 (COVID‐19).1 More than fifty long term effects have been documented, including dyspnoea, fatigue, cognitive changes, and altered mood.2 A United Kingdom multicentre cohort study (1077 patients) found that almost one in two people hospitalised with COVID‐19 did not consider themselves fully recovered twelve months after discharge.3

The World Health Organization defines the post‐COVID‐19 condition (“long COVID”) as the persistence or new development of symptoms, typically three months after infection with the severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2), that last for at least two months and are not explained by an alternative diagnosis.4 The condition can cause substantial disability, hinder social participation, and require continuing health care. There is no consensus regarding the best model of care for people with long COVID, and no specific treatments. United Kingdom guidelines recommend excluding other medical causes for the symptoms, optimising management of other medical conditions, providing multidisciplinary rehabilitation, and supporting the individual manage their condition and return to work or study.5

The Alfred Health Post‐COVID service was established in June 2020 as a triage and referral service for people with persistent symptoms. Its aim is to provide specialist assessment and management for people with complex needs, and to return them to community‐based care once the required services are available. The multidisciplinary team at the service includes general medicine physicians, physiotherapists, allied health assistants, and neuropsychologists; patients can also be referred to other medical, rehabilitation, and allied health services as needed.

In this article, we describe the operation of the Alfred Health Post‐COVID service during its first three years, with the aim of providing information for informing the design of similar services. We report the proportion of people who had been hospitalised with COVID‐19 or referred to the service by general practitioners who completed the Alfred Health Post‐COVID service screening questionnaire, the characteristics of people who completed the questionnaire, the characteristics of respondents who had not returned to work or study after COVID‐19, and the specialist services required by people with persistent symptoms after COVID‐19.

Methods

In our observational cohort study, all people admitted to Alfred Health with a diagnosis of COVID‐19 — both inpatients and those managed as hospital‐in‐the‐home patients (a home‐based acute medical service) — were invited by text message or telephone call approximately eight weeks after their initial admission to hospital to register for follow‐up assessment. During 1 January – 31 December 2022, people with persistent symptoms could also be referred by general practitioners in the Alfred Health catchment area. Alfred Health is a major tertiary referral hospital, with more than 600 acute care beds, in inner southeastern Melbourne.

People who accepted the invitation could complete symptom questionnaires managed with Research Electronic Data Capture (REDCap), hosted at Alfred Health; for people without access to REDCap, questionnaires were administered by telephone. The questionnaire screened for symptoms often reported after COVID‐19:2 dyspnoea (Modified Medical Research Council [mMRC] scale6), fatigue (Fatigue Assessment Scale, FAS7), anxiety and depression (Hospital Anxiety and Depression Scale, HADS8), and distress related to the experience of COVID‐19 (ie, symptoms of post‐traumatic stress disorder [PTSD]; Impact of Events Scale–Revised, IES‐R9). General health status and health‐related quality of life were assessed with the EQ‐5D‐5L.10 Questions about returning to work, weight loss, and altered cognition and memory were also included (Supporting Information, part 1). Finally, a free text question enabled people to report symptoms not otherwise covered by the questionnaire.

We report data for people admitted to hospital with COVID‐19 (or who had COVID‐19 during this period and were later referred to our service by their general practitioners) during 19 March 2020 – 28 December 2022 who completed 8‐week follow‐up assessments during 3 June 2020 – 3 March 2023; some people also completed 12‐month follow‐up assessments during 28 June 2021 – 3 March 2023. We extracted the data for our analysis on 14 April 2023.

Further care

A physiotherapist reviewed the responses to the screening questionnaire and discussed care pathways with each respondent. People with complex symptoms, comorbid conditions, fatigue, or altered cognition and memory were assessed in the general medicine service. People could also be referred to other specialist medical services at Alfred Health (eg, respiratory medicine, cardiology, neurology) when indicated. People with new functional limitations were assessed in the physiotherapy service, including functional exercise testing when required, and could be referred to rehabilitation services (eg, pulmonary rehabilitation). Those with altered memory or cognition (eg, brain fog) were offered neuropsychology assessment and care. People who required psychological support were encouraged to consult their general practitioners about a mental health treatment plan for care in the community. Each patient received information about self‐management of symptoms often experienced by people after COVID‐19.11 A fortnightly multidisciplinary team meeting (general medicine physician, physiotherapist, allied health assistant, neuropsychologist) discussed the management of patients with complex needs. People were discharged to ongoing community care once the required services were available. People were invited to complete the same REDCap questionnaire twelve months later to evaluate the persistence of symptoms.

Statistical analysis

As Victoria experienced three distinct COVID‐19 waves with different dominant SARS‐CoV‐2 variants, we report results by year of acute COVID‐19. We report the numbers and proportions of eligible people who elected to complete follow‐up screening at eight weeks (uptake). We report the numbers and proportions of people with persistent symptoms (based on assessment instrument scores) according to the usual cut‐off values for these instruments. For health status, we report the number and proportion of people with at least moderate limitation (score of 3 or more) in the corresponding EQ‐5D‐5L domain. We report the numbers and proportions of participants who required specialist medical, allied health, or rehabilitation services. Proportions for each symptom and characteristic exclude non‐responses to the corresponding question; that is, lack of response was not interpreted as the symptom or characteristic being absent.

We assessed the statistical significance of differences in the proportions of people who had or had not returned to work or study by the 8‐week follow‐up who reported selected symptoms in χ² or independent sample t tests. Symptom variables for which the differences were statistically significant were included in a binary logistic regression model, using backward variable selection. To avoid multicollinearity, only one symptom and one health‐related quality of life domain were included in the final multivariable binary logistic regression model; we report adjusted odds ratios (aORs) with 95% confidence intervals (CIs). All analyses were undertaken in SPSS 28. P < 0.05 was deemed statistically significant.

Ethics approval

The study was approved by the Alfred Hospital Human Research Ethics Committee (HREC/16/Alfred/59). As the study was a service evaluation, the committee waived the requirement for individual informed consent.

Results

During 19 March 2020 – 28 December 2022, 6712 eligible people were invited eight weeks after the onset of acute COVID‐19 to complete symptom questionnaires; 1124 accepted the invitation (17%) and 726 completed the questionnaire (11%) (Box 1). The mean age of people who completed the questionnaire was 53 years (standard deviation [SD], 15 years; range, 18–78 years); 473 were women (65%). Most respondents (721, 99%) had experienced COVID‐19 only once; five participants reported (in 2022) having had COVID‐19 twice. In 2020, 24 of 25 people who accepted the invitation to participate (96%) had been hospitalised with COVID‐19, and six of 21 (29%) had been admitted to intensive care; in 2022, 136 of 202 people who accepted the invitation (67%) had been hospitalised and three of 194 (1%) had been admitted to intensive care (Box 2).

Eight‐week survey

Eight weeks after the onset of COVID‐19, 385 of 642 respondents reported at least one persistent symptom (60% of respondents to the symptom questions, 5.7% of all invitees). The proportions who reported dyspnoea (five of 13 respondents, 39%) or extreme fatigue (four of eleven, 36%) were largest in 2020; in 2021 and 2022, 371 of 656 respondents (57%) reported memory changes and 431 of 656 concentration changes (66%); neither symptom had been assessed in 2020. Across the three years, mobility limitations were reported by 162 of 688 respondents (24%) and self‐care limitations by 41 of 691 (6%); 253 of 687 (37%) reported pain and discomfort that affected health‐related quality of life, and 257 of 689 (37%) limitations to usual activities (Box 2). Symptom prevalence was similar for people who were admitted to intensive care or as hospital inpatients or hospital‐in‐the‐home patients; the proportions who reported symptoms were larger among people referred by general practitioners because of persistent symptoms (Supporting Information, table 1).

Free text responses

The free text responses provided by 442 of the 726 respondents (61%) described a broad range of further persistent symptoms, including loss of taste and smell, tinnitus, headaches, physical pain, gastrointestinal symptoms, and hair loss. Many respondents referred to brain fog and its impact on work and study, and to exacerbation of other conditions, such as asthma (Box 3). Fears for the future and frustration with the lack of diagnostic tests for long COVID were often expressed.

Return to work or study after COVID‐19

Sixty‐seven of 453 respondents who had been working or studying prior to COVID‐19 (15%) had not returned to these activities by the 8‐week follow‐up. The proportions of respondents who reported dyspnoea (56% v 25%), depression (48% v 19%), fatigue (69 v 27%), or distress (47% v 21%) were larger than for people who had not returned to work or study than for those who had, and their median health‐related quality of life score was lower (EQ‐5D‐5L visual analogue scale (34; interquartile range [IQR], 3–95 v 65; IQR, 4–100) (Supporting Information, table 2).

Our regression analysis included data for the 421 respondents for whom complete data were available regarding dyspnoea, extreme fatigue, depression, mobility limitation, and memory changes. People who reported extreme fatigue (aOR, 0.25; 95% CI, 0.13–0.46) or mobility limitations (aOR, 0.29; 95% CI, 0.16–0.52) were less likely to have returned to work or study than those who did not (Box 4).

Further care

Of the 726 questionnaire respondents, 660 (91%) were assessed as requiring further care, which was coordinated by the Post‐COVID service. The specialist medical services most frequently required were general medicine (132 people, 18%) and respiratory medicine (92, 13%). A smaller number of referrals were made to specialist clinics for cardiology, neurology, haematology, ear nose and throat, sleep and dermatology. All patients were encouraged to see their general practitioners for ongoing care, including mental health care plans for psychological support. People also received a range of allied health services, most frequently physiotherapy (99 people, 14%), neuropsychology (92, 13%), and pulmonary rehabilitation (33, 5%). Eight people were referred for occupational therapy (including six for vocational rehabilitation), two each to pain specialists, speech pathology, and dietetics, and one each to a diabetes educator and community rehabilitation (Box 5).

Twelve‐month follow‐up survey

Seventy‐one people also completed follow‐up questionnaires at about twelve months (mean, 386 days after COVID‐19; 10% response rate with respect to the 8‐week survey); their mean age was 52 years (range, 23–77 years), 42 were women (59%), 47 had been admitted to hospital (66%), and five had been admitted to intensive care (7%). Three respondents reported having again been ill with COVID‐19 since their first contact with the service. The proportions of respondents to the 12‐month questionnaire who had reported persistent symptoms at eight weeks were larger than for all respondents at eight weeks. For those who completed both surveys, the proportions who reported dyspnoea, depression, or mobility limitations were smaller at twelve months than at eight weeks; the proportions who reported persistent extreme fatigue, anxiety, or memory and concentration changes were similar at both time points. The proportion of respondents to both surveys who had not returned to work or study was similar at eight weeks (nine, 23%) and twelve months (eleven, 28%) (Box 6), but was larger than the proportion among all eight‐week respondents (67 of 453, 15%).

Discussion

We have described the experience of one of the longest running post‐COVID‐19 services in Australia. Of 6712 people hospitalised at Alfred Health with COVID‐19 during 2020–2022 or referred by their general practitioners during 2022, 17% elected to receive specialist follow‐up at eight weeks, and 11% completed screening questionnaires. Sixty percent of people who provided information about symptoms reported at least one persistent symptom at eight weeks, most frequently altered memory and concentration, dyspnoea, or fatigue. Fifteen percent of those previously in employment or study had not returned to it eight weeks after their acute illness. Ongoing support required a range of medical and non‐medical services. Many of the small proportion of people who completed follow‐up surveys at twelve months reported persistent fatigue, anxiety, and cognitive changes.

The proportion of people with COVID‐19 who later develop persistent symptoms, including long COVID, is unknown. The wide variation in reported estimates may reflect differences between patient cohorts, definitions of long COVID, and epidemiological changes during the pandemic, including COVID‐19 vaccination rates.12 Eight weeks after acute COVID‐19, persistent symptoms were reported by 60% of survey participants who provided relevant information, or 5.7% of people eligible for the service. The estimated uptake of our follow‐up service is consistent with the estimated prevalence in the United Kingdom of post‐COVID‐19 symptoms that affect daily life (about 5%).1

Our findings may be useful for planning health care services for people with long COVID. Consistent with overseas reports,3 we found no relationship between the persistence of symptoms and severity of the acute illness (as indicated by admission to intensive care), a problem for identifying those who might require longer term follow‐up. Our approach to post‐COVID‐19 care provided people the opportunity with persistent symptoms to receive appropriate care at eight weeks; we selected this time point, earlier than the 12‐week time point typically applied for diagnosing long COVID,5 to facilitate earlier intervention.

Long COVID is complex, affecting several body systems,2 and it is likely that not all manifestations can be managed at a single clinic. Our model focused on comprehensive screening and assessment, and many people were referred to other services for further care. The allied health‐led model of care offered screening to more than 6000 people; of the 726 people who used the service for 8‐week follow‐up, only one‐third required specialist medical review. The Alfred Health Post‐COVID service provided access to care that would otherwise not be readily accessible in the health system or the community, including dedicated general medicine assessment for people with persistent symptoms, and neuropsychological assessment and management. Ongoing management by general practitioners was important for all patients. Service gaps remained; for example, 67 people had not returned to their previous work or study, but only eight had been referred for vocational rehabilitation. Referral pathways might be different in areas where different clinical services are available. Research is needed to determine the optimal model of assessment and follow‐up for people with long COVID.

Limitations

Strengths of this study were our use of validated instruments to assess the presence of persistent symptoms, and we invited all people with COVID‐19 managed by our health service for follow‐up, which allowed us to accurately report demand for our service amongst admitted patients. We identified a small group of people with complex needs who required specialist services. Although our service was established to provide care for people who had been admitted to hospital with COVID‐19, 13% of respondents who had COVID‐19 during 2022 had been managed in the community. It is not clear whether they required different services to people who had been hospitalised, but the prevalence and nature of post‐COVID‐19 symptoms are similar for hospitalised and non‐hospitalised people.13 We could not distinguish between new symptoms and those attributable to other health conditions. Our study did not include a contemporary control group, of particular relevance given the high population prevalence of mood disorders during the COVID‐19 pandemic.14 We had no information about the time taken to return to work by those who did so. As only a small number completed the 12‐month follow‐up, the impact of our service is difficult to assess. In the absence of specific treatments for long COVID,12 we were limited to providing thorough assessment, optimal management of other medical conditions, rehabilitation, and supportive care in accordance with guidelines for managing long COVID.5 If specific treatments for long COVID become available, the structure of such services could change considerably.

Conclusion

Our post‐COVID‐19 service model, including voluntary use and systematic screening, found that 5.7% of eligible people had persistent symptoms eight weeks after developing acute COVID‐19. Management and onward referral involved a variety of clinical services, most frequently general medicine, respiratory medicine, physiotherapy, neuropsychology, and pulmonary rehabilitation. Possible shortcomings in care included vocational rehabilitation for people who had not returned to work. Our findings may assist the planning of services for people with long COVID.