Biochemical markers of severity and outcome in the older patient with Covid-19

Background

The difference in outcomes of mortality observed across different age groups has been a major influence in the management of Covid-19. Not only is there a higher rate of mortality in the older population, but poorer outcomes and severe presentations are also particularly evident in older people.^1,2 In a study of 20,133 UK patients, the median age of patients who died was 80 years.¹

The initial clinical picture of Covid-19 was one of respiratory disease. It soon became apparent that there was a differing clinical presentation, particularly in older and frail adults. This variability of presentation has meant that objective measures are helpful in the forming of a diagnosis of Covid-19 and to prognosticate disease severity and mortality.

D-dimer, ferritin, and troponin are biochemical markers that have been identified to assist in making a diagnosis and predicting clinical outcome. Normal physiological ranges are determined by accepting results that occur within 95% of a fairly homogenous, fit population.³ The clinical question here is, is this appropriate in the frail older population? Given the disparity in presentation, outcomes and mortality in the older population, the question arises as to whether the use of these markers as prognosticators of outcome remains valid and if so, what serum concentration is significant.

Our study consisted of two different cohorts, patients under the age of 75 and patients who were 75 and older, with one hundred patients in each cohort. We sex matched each patient such that there were fifty men and fifty women in each cohort.

The aims of this study are to:

1. Compare serum levels of biochemical markers.
2. Ascertain if there are differences in clinical outcomes in older and younger Covid-19 patients.
3. Identify rate of readmission, place of discharge and Covid-19 status in older patients.
4. Assess the ability of the Clinical Frailty Scale (CFS) to predict mortality in Covid-19 patients.⁴

Methods

Patients admitted with a confirmed diagnosis of Covid-19 were identified from the electronic patient record. Diagnosis of Covid-19 was either on the basis of a positive SARS-CoV-2 polymerase chain reaction test on a nasopharyngeal swab, or respiratory sample. Patients who had clinical signs, symptoms and supportive investigations of Covid-19 as assessed by the lead medical clinician managing the patient, but who had no positive microbiological diagnosis of Covid-19 were also included and are referred to as clinically diagnosed patients henceforth.

This was a prospective cohort study, conducted on patients admitted between 2nd April 2020 and 23rd April 2020. Data was collected at three time points, admission, discharge, and at 10 months. Patients were identified by two of the authors reviewing the lists of patients admitted within the preceding 24 hours. 200 patients were identified; 100 of 75 years of age and over and these were sex matched 1:1 to controls who were 74 years of age and younger.

Patient baseline demographics, as well as admission dependency, admission frailty (measured by CFS, as recommended by the National Institute for Health and Care Excellence), initial serum D-dimer, ferritin and troponin were collected.⁵

Data collected on discharge included CFS, mortality, discharge destination, and dependency. Dependency at the time of admission and at discharge were stratified into seven distinct groups; alone and independent, alone with a package of care, with family, with family and with a package of care, residential home, nursing home and rehabilitation unit. Data was collected at admission, discharge, and at 10 months to assess mortality and readmission rates. If readmitted, their frailty and dependency was noted. Collected data was analysed using Microsoft Excel Office 2019 and IBM SPSS Statistics version 26.

Ethical approval

This study made use of information collected as part of routine care, and therefore did not require formal ethics committee approval. Approval was sought and granted by the local clinical effectiveness department.

Demographics

100 patients aged ≥75 (mean age 84.5 years, range 75 to 98 years) were compared to 100 patients aged under 75 years (mean age 56.9 years, range 26 to 74 years), 49% were male. Those ≥75 year had a median and mode admission frailty score of 6, and most (38%) lived with family but had no care package prior to admission. 21% lived alone without a care package, 16% lived in a nursing home, 12% lived alone with a care package, 9% lived in a residential home, 7% lived with family but had a care package, 1% were in a rehabilitation unit, and for 3% lacked data on their living situation.

Diagnosis of Covid-19

Those ≥75 years and those <75 years were equally likely to be diagnosed with Covid-19 by a positive respiratory viral swab than clinical impression (79% vs 81%; p=0.724).

Laboratory markers

Those ≥75 years had slightly lower initial mean serum ferritin, but this was not statistically significant (1352ug/L vs 1476ug/L, p=0.896). Older patients also had slightly lower mean D-dimers, but this was also not statistically significant (3188ng/L vs 3365ng/L, p=0.370). A logistic regression was performed to ascertain the effects of biochemical markers on the likelihood that patients died. The model explained 26.7% (Nagelkerke R2) of the variance in mortality and correctly classified 76.6% of deaths. D-dimer was found to contribute to the model and was associated with mortality (p=0.039). Ferritin and troponin did not predict mortality.

Older patients have significantly higher mean serum troponins (88.1ng/L vs 35.9, p=0.007).

Venous thromboembolism (VTE)

There was significantly less investigation for deep vein thrombosis (DVT) and pulmonary emboli (PE) in older patients with (9% vs 18%, p= 0.030), and fewer diagnoses made (and 6% vs 8%, p= 0.579). There was a non-significant pattern noted in those investigated for VTE; older patients investigated for VTE were more likely to have VTE compared to younger Covid patients investigated for VTE (66.7% vs 44.4%, p=0.276).

Another nonsignificant trend noted, was that in Covid-19 patients with PE, older patients had lower D-dimers than younger controls (3706ng/L vs 4385ng/L, p=0.275).

Mortality and dependency on discharge

Mortality during admission was greater in the older age group, (55.0% vs 25.0%, with an odds ratio of 3.67 p<0.0001). Age ≥75 years positively predicted mortality (p=0.005). Male sex was associated with a 49% increase in mortality (OR 1.49, 95% confidence interval = 0.846 to 2.638).

12.9% of patients≥75 years who were discharged were more dependent and required a change to their discharge support. A logistic regression was performed to ascertain the effects of frailty on the likelihood that patients required increased support on discharge. The model explained 50.6% (Nagelkerke R2) of the variance in support on discharge and correctly classified 86.4% of patients needing increased support on discharge. Laboratory markers did not predict an increase in support on discharge (p=0.404).

Similarly, a logistic regression was performed to ascertain the effects of frailty on the likelihood that patients died. The model explained 54.0% (Nagelkerke R2) of the variance in mortality and correctly classified 76.0% of patients dying.

10-month follow up

	≥75 years	Under 75 years
Readmission within 10 months n (%)	15 (33.3%)	6 (8.0%)	p<0.001
Number of readmissions mean (range)	0.73 (0 to 8)	0.09 (0 to 1)	p=0.007
Mortality at 10 months following initial discharge	3 (6.7%)	0 (0%)	p<0.001

Table 1: Readmission and mortality of patients at 10 months post initial admission with COVID-19

15 out of 45 (33.3%) surviving older patients were readmitted within 10 months of their initial admission, compared to just 6 out of the surviving 75 (8.0%) younger patients (p<0.001, Table 1). Mean number of readmissions in over 75 years was 0.73 but ranged from 0 to 8 readmissions. This was compared to a mean readmission rate of 0.09 for controls under 75 years, p=0.007. Mortality remained high in the older group at 10 months, 3 (6.7%) of those ≥75years who survived their initial admission with Covid-19 died in the subsequent 10 months, whereas none of their younger counterparts did (p<0.001, Table 1). For those that survived and were subsequently readmitted, they tended to become more frail with subsequent admissions, but stabilise over second and third readmissions.

For those that survived and were subsequently readmitted, their dependency on discharge often increased, but on subsequent admission remained static. There was no statistically significant relationship between initial admission frailty and number of readmissions in the following 10 months, however readmissions tended to increase with increasing frailty. On first admission, there were most deaths in those who were more frail (CFS 6 and 7), however deaths in the following 10 months were in those less frail (CFS 1, 3 and 4).

Discussion

Numerous algorithms and methodologies have been suggested on how to best prognosticate patients with Covid-19 and stratify according to severity in the face of an overstretched healthcare system.

We have investigated the role of biochemical markers as a possible prognosticator for mortality and whether these results would differ among older and younger patients. Our results showed serum troponin was significantly raised in older patients when compared to a younger group. Similar results have been observed in other studies which showed patients with raised troponin were more likely to be older with more comorbidities and worse prognosis.^6,7

However, a raised troponin is frequently observed in other conditions such as severe pneumonia, sepsis, critically unwell patients often independent of an acute coronary syndrome.⁸ A raised troponin did not significantly predict mortality in the older group, and this has also been shown in other studies.⁹ We can conclude that there are many factors which would contribute to raised troponin independent of Covid-19. This suggests a common causal pathway in different clinical presentations leading to a raised troponin in older patients.

D-dimer was the only biomarker that was significantly associated with mortality and this was noted in both cohorts of patients, perhaps a marker of severity. Yao et al. demonstrated that D-dimer predicted mortality and also severity of Covid-19 disease.¹⁰ It is important to note that in our study, there were other confounders that may affect mortality rates and are associated with a raised D-dimer including VTE and PE, inflammation or DIC (Disseminated Intravascular Coagulation).

The prevalence of PE in particular is well described and also anecdotally described in Covid-19.¹¹ Within our study, 7% of patients admitted were diagnosed with thromboembolic events. Studies have suggested a high D-dimer should lead to a low threshold for investigations for VTE.

Khan et al suggest a D-dimer >2000ng/ml should indicate a low threshold for investigation for VTE in Covid-19.¹² In our study only 9% of older patients were investigated for DVT, compared to 18% in their younger counterparts; despite both cohorts having elevated mean D-dimers. The reason for this under-investigation in our sample is unclear. Some propositions we offer are, the older group had a slightly lower mean D-dimer compared to the younger patients although this wasn’t statistically significant. This could also be related to the generally low admission rates of older patients to higher level care and as such fewer investigations performed.

Although there were less investigations for VTE in the older cohort, those that were investigated were more likely to have a positive diagnosis. This could be related to better clinical judgement or more judicious use of investigations by geriatricians. Also given the severity and high mortality rate of Covid-19 in older patients, ultrasound dopplers or computed tomography scanning may have been thought not to change outcome in some patients. As with the offer of cardiopulmonary resuscitation and mechanical ventilation, bias, potential harms and fairness come in to play.¹³

In our patient cohort, there was a high mortality, with those ≥75 years significantly more likely to die during admission (55.0% vs 25.0%, p<.0001) with an odds ratio of 3.67. This is in keeping with a Public Health England report that showed that patients in a survival analysis aged 80 and older were seventy times more likely to die compared to patients under the age of 40.¹⁴

Our mortality rate could also be affected by the period of data collection – as a Lancet study demonstrated that in-hospital UK mortality improved over time from 52% in the first week of March to 17% in the last week of May, likely related to hospital strain and changes in practice as experience and knowledge were acrrued.¹⁵ The higher mortality in the older population relates to increased comorbidities, frailty and diminishing physiological reserve, and this is important given the ageing population of the UK.

Age may not be directly related to frailty and therefore risk of deterioration still needs to be assessed on an individual patient by patient basis, using the CFS to aid this. NICE guidelines recommend the use of CFS in decisions in management of Covid-19 including admission to critical care as frailty is analogous to premorbid functional status. In addition, as expected, we found that male sex predicted mortality. The increased risk of death in the male sex is consistent with UK hospital data.¹⁵

The older UK population is growing rapidly, with associated comorbidities, frailty and dependency. Our study also reviewed secondary outcomes in the older cohort on admission and also at 10 months follow up. We reviewed the impact of Covid-19 on discharge destinations, discharge frailty scores and readmission to the hospital. A review of discharge destinations provided an indirect measurement of the level of support a patient received on discharge and the impact of Covid-19 on their quality of life. Only 12.9% of the patients ≥75 years had a change to their discharge destination. This implies that although older patients were more likely to have poorer outcomes after Covid-19, they mostly returned to baseline on discharge.

Of note, in those who were subsequently readmitted, their discharge dependency often increased on the next admission but then remained static. Potentially, this could mean other illnesses leading to subsequent admissions are more likely to lead to increased discharge dependency compared to Covid-19; however more information would be required to fully evaluate this association.

This may also be due to the fact that those who are severely frail may already be receiving maximal care in the community. This could be elicited with a larger study, as only 45 patients over 75 from the original 100 patient cohort survived for inclusion in readmission statistics.

At 10 months follow up, 6.7% of patients ≥75 years had died and none of the younger group had passed away. Older patients were significantly more likely to be readmitted in the 10 months than the younger cohort. Given that CFS is a recommended score as per NICE guidelines in decision making and management of Covid-19, we reviewed the impact of frailty scores on outcomes in our older cohort. On first admission with Covid-19, there were most deaths in those with a higher admission frailty (6 and 7) however our study did not show a statistically significant correlation between the two.

Neither did frailty predict discharge dependency in those who survived Covid-19, also described in other studies.^16,17 We observed a step down of frailty scalse on the first readmission which then stabilises over the next 10 months. This underlines the importance of discussions around escalation and discharge planning with older patients and their family members on their initial admission with Covid-19, as we know on subsequent admissions they will be more frail and dependent. Frailty is also an indicator of outcome following cardiopulmonary resuscitation and admission to intensive care, therefore a useful scale for the basis of these discussions.^18,19

Strengths and weaknesses

A strength of this study was the duration of follow up; 10 months, which enabled us to establish post-Covid-19 mortality, and survival through the second wave of the pandemic. In addition, our controls were sex matched 1:1 within our local population.

A weakness of this study was that timing of initial Covid-19 blood tests (ferritin, troponin and D-dimer) were not consistent across patients. Some patients had their Covid-19 blood tests on the day of admission, whereas others on the day after or not at all. It appeared older patients were less likely to have Covid-19 blood tests done. As a result of this, there were data gaps which may have reduced the statistical significance of our results.

20% of patients in our study diagnosed with Covid-19 did not have a positive Covid-19 swab. As we included those clinically diagnosed as well as laboratory-diagnosed, our results may have included some patients who did not have true Covid-19 but a similar clinical picture. At this stage in the first wave many patients were diagnosed clinically due to the delay, inaccuracy and availability of swab results.

With 100 patients and 100 controls, our work could be improved with a larger sample size. In addition, due to the high mortality rate, fewer patients could be included in analysis regarding discharge dependency and readmission.

Variability in frailty may have arisen as frailty was assessed according to documentation in the medical notes. If a patient did not have a CFS documented, it was interpreted from their medical and social history. To counter any inter-observer variability, those involved in the study attended the same teaching session by the lead Geriatrics Consultant. In addition, a recent study demonstrated high inter-observer reliability and precision for retrospective use of CFS in acute care.²⁰

There may also have been variability of function within patients discharged to a rehabilitation unit leading to heterogeneity in patients classified in this group.

Conclusions

We found significant differences between older and younger patients with Covid-19. Patients ≥75 years had significantly higher admission troponins than younger patients, in keeping with other studies and need not necessitate further investigation or intervention unless there is specific clinical concern.

They are less likely to be investigated for VTE however if investigated, more likely to be diagnosed with a PE or DVT. D-dimer and age ≥75 as well as male sex were associated with mortality, and therefore these criteria should be used to identify patients at risk of deterioration. There appeared to be more deaths on first admission with more frail patients, however only a small proportion of patients ≥75years that survived required a change to their discharge destination.

These older patients who survived and were discharged had an increase in their dependency on discharge after the first readmission, but remained stable following their second and third readmissions in the following 10 months. This underlines the importance of discussions around frailty, escalation and discharge planning with older patients with Covid-19 pneumonitis and their family members on their initial admission.

---

Dr Penelope Teoh, ST3 Specialist Registrar in Microbiology and Infectious Diseases, University College London Hospitals NHS Foundation Trust

Abisoye Olamide Akintimehin, Queen Elizabeth Hospital, Lewisham & Greenwich NHS Foundation Trust

David Smithard, Queen Elizabeth Hospital, Lewisham & Greenwich NHS Foundation Trust and University of Greenwich

---

References

1. Docherty AB, Harrison EM, Green CA, et al. Features of 20 133 UK patients in hospital with covid-19 using the ISARIC WHO Clinical Characterisation Protocol: prospective observational cohort study. BMJ 2020;369:m1985. DOI: 10.1136/bmj.m1985.
2. Landi F, Barillaro C, Bellieni A, et al. The New Challenge of Geriatrics: Saving Frail Older People from the SARS-COV-2 Pandemic Infection. J Nutr Health Aging 2020;24(5):466-470. DOI: 10.1007/s12603-020-1356-x.
3. Sikaris KA. Physiology and its importance for reference intervals. Clin Biochem Rev 2014;35(1):3-14. (https://www.ncbi.nlm.nih.gov/pubmed/24659833).
4. Church S, Rogers E, Rockwood K, Theou O. A scoping review of the Clinical Frailty Scale. BMC Geriatr 2020;20(1):393. DOI: 10.1186/s12877-020-01801-7.
5. NICE. COVID-19 rapid guideline: critical care in adults. NICE guideline [NG159]. London. 2021.
6. Tersalvi G, Vicenzi M, Calabretta D, Biasco L, Pedrazzini G, Winterton D. Elevated Troponin in Patients With Coronavirus Disease 2019: Possible Mechanisms. J Card Fail 2020;26(6):470-475. DOI: 10.1016/j.cardfail.2020.04.009.
7. Webb IG, Yam ST, Cooke R, Aitken A, Larsen PD, Harding SA. Elevated baseline cardiac troponin levels in the elderly – another variable to consider? Heart Lung Circ 2015;24(2):142-8. DOI: 10.1016/j.hlc.2014.07.071.
8. Lee YJ, Lee H, Park JS, et al. Cardiac troponin I as a prognostic factor in critically ill pneumonia patients in the absence of acute coronary syndrome. J Crit Care 2015;30(2):390-4. DOI: 10.1016/j.jcrc.2014.12.001.
9. Brill SE, Jarvis HC, Ozcan E, et al. COVID-19: a retrospective cohort study with focus on the over-80s and hospital-onset disease. BMC Med 2020;18(1):194. DOI: 10.1186/s12916-020-01665-z.
10. Yao Y, Cao J, Wang Q, et al. D-dimer as a biomarker for disease severity and mortality in COVID-19 patients: a case control study. J Intensive Care 2020;8:49. DOI: 10.1186/s40560-020-00466-z.
11. Jevnikar M, Sanchez O, Humbert M, Parent F. Prevalence of pulmonary embolism in patients with COVID-19 at the time of hospital admission and role for pre-test probability scores and home treatment? Eur Respir J 2021. DOI: 10.1183/13993003.01033-2021.
12. Khan MZ, Jamal Y, Sutton B, Rauf F. Venous thromboembolism in patients with COVID-19 and correlation with D-dimers: a single-centre experience. BMJ Open Respir Res 2020;7(1). DOI: 10.1136/bmjresp-2020-000779.
13. Smithard DG, Abdelhameed N, Han T, Pieris A. Age, Frailty, Resuscitation and Intensive Care: With Reference to COVID-19. Geriatrics (Basel) 2021;6(2). DOI: 10.3390/geriatrics6020036.
14. England PH. Disparities in the risk and outcomes of COVID-19. PHE publications. 2020. (GW-1447).
15. Navaratnam AV, Gray WK, Day J, Wendon J, Briggs TWR. Patient factors and temporal trends associated with COVID-19 in-hospital mortality in England: an observational study using administrative data. Lancet Respir Med 2021;9(4):397-406. DOI: 10.1016/S2213-2600(20)30579-8.
16. Miles A, Webb TE, McLoughlin BC, et al. Outcomes from COVID-19 across the range of frailty: excess mortality in fitter older people. Eur Geriatr Med 2020;11(5):851-855. DOI: 10.1007/s41999-020-00354-7.
17. Owen RK, Conroy SP, Taub N, et al. Comparing associations between frailty and mortality in hospitalised older adults with or without COVID-19 infection: a retrospective observational study using electronic health records. Age Ageing 2021;50(2):307-316. DOI: 10.1093/ageing/afaa167.
18. Muessig JM, Nia AM, Masyuk M, et al. Clinical Frailty Scale (CFS) reliably stratifies octogenarians in German ICUs: a multicentre prospective cohort study. BMC Geriatr 2018;18(1):162. DOI: 10.1186/s12877-018-0847-7.
19. Lopez Cuenca S, Oteiza Lopez L, Lazaro Martin N, et al. Frailty in patients over 65 years of age admitted to Intensive Care Units (FRAIL-ICU). Med Intensiva (Engl Ed) 2019;43(7):395-401. DOI: 10.1016/j.medin.2019.01.010.
20. Stille K, Temmel N, Hepp J, Herget-Rosenthal S. Validation of the Clinical Frailty Scale for retrospective use in acute care. Eur Geriatr Med 2020;11(6):1009-1015. DOI: 10.1007/s41999-020-00370-7.