COVID-19 pneumonia in older patients: clinical features and outcomes of a 400 case series

Introduction

In late 2019, a novel human coronavirus named severe acute respiratory syndrome coronavirus 2 (SARS-Cov-2) was firstly identified in Wuhan, China, as the cause of coronavirus disease 2019 (COVID-19).¹ Ever since, COVID-19 out-break became a major health issue worldwide.

COVID-19 presents a wide spectrum of symptoms from asymptomatic forms with sometimes loss of taste and smell,² mild symptoms including lethargy, fever and cough to severe hypoxemic pneumonia up to acute respiratory distress and multiorgan failure.

It was clear that age and comorbidities are major prognostic factors within the literature.³ Indeed, age has been strongly associated with poor outcomes and from the earliest stages of the pandemic, age dependent risk of infection and mortality was reported by several studies.^4,5

The rate of older patients in study cases represented, among infected patients, from 15.1% in Guan et al. study,⁶ up to 52% in Yang et al. study.⁷

Older patients seem more susceptible to severe forms and more prone to severe outcomes with higher mortality rates. Due to frailty and multimorbidity, they require longer hospital stays and more assistance.⁸

This susceptibility is due to physiological changes such as anatomic changes in nasal cavity volume,⁹ mucociliary clearance time¹⁰ and reductions in lung physiological capacity¹¹ exposing elderly to infections of the upper and lower respiratory tract. Frailty and respiratory impairment are highly associated and when both are present the death risk increases substantially.¹² This susceptibility is also due to immunosenescence, defined as immune system decline¹³ and age-related decline in cellular repair mechanisms that causes increase pro-inflammatory cytokine production resulting in “inflammaging”.¹⁴

Clinical presentation in older patients may be challenging since it can be atypical with absence of fever and absence of respiratory symptoms yet with exacerbation of heart failure and neurological symptoms such as delirium and confusion or strokes.¹⁵ These factors contribute to heavier charges on the healthcare system.

The aims of this study were to describe the clinical features, biological and radiological abnormalities in older patients with SARS-Cov-2 pneumonia, to report complications and mortality rate among the study group and to identify factors associated with poor outcomes.

Methods

Results

Descriptive analysis

Demographic data

The study group included 222 women (55.5%) and 178 men (Figure 1).⁵⁹ The mean age was 74 years [65-97 years] and 77% of patients (N=308) were aged between 65 and 79.

Figure 1. Flow chart of the inclusion process.

Co-morbidities

In total, 30% of the study population (N=120) had more than two co-morbidities while 27% (N=108) had two co-morbidities and 27.75% (N=111) had just one co-morbidity. So, in total 84.75% of the study group had at least one chronic disease. The most frequent chronic disease was high blood pressure (55.75%) whether it was known before admission in 203 cases (91%) or discovered during the hospital stay in 10 cases (9%). The second most frequent co-morbidity was diabetes mellitus (50%); In 24% of cases (N=48), it was discovered upon admission. 48% of patients (N=96) were not within treatment target and 22.5% (N=45) had chronic complications such as coronary arteries disease in six patients(13.33%), diabetic nephropathy in 26 cases (57.78%), neuropathy in four patients (8.89%), retinopathy in six cases (13.33%), feet problems in three patients (6.67%). Pre-diabetes was found in 18% of cases (N=72). The third most frequent comorbidity was hyperlipemia (25.25%) followed by obesity in25% of cases.

Clinical presentation

The most common symptoms were dyspnoea (69.8%), fatigue (66.8%), and dry cough (65.8%).

Electrocardiogram (ECG) abnormalities were the most common physical examination abnormalities observed in the study including corrected QT (cQT) interval prolongation, observed in 97 cases (24.2%) on admission. Among them, 60 ECGs (61.8%) were controlled revealing that 39 (65%) returned to normal range.

Among the 39 patients who had prolonged corrected QT interval that returned to normal, 24 (61.5%) were exposed to azithromycin before admission.

Clinical features are presented in Table 1.

Table 1. Clinical features of SARS-Cov-2 pneumonia.

Clinical features	Number of patients	Percentage %
Symptoms
Dyspnoae	279	69.8%
Fatigue	267	66.8%
Dry cough	263	65.8%
Fever	222	55.5%
Diarrhoea	110	27.5%
Vomiting	89	22.3%
Headache	83	20.8%
Anorexia	78	19.6%
Productive cough	55	13.8%
Anosmia	53	13.3%
Ageusia	51	12.8%
Myalgia	44	11%
Nausea	27	6.8%
Weight loss	26	6.5%
Abdominal pain	22	5.5%
Chest pain	15	3.8%
Arthralgia	5	1.3%
Physical examination parameters
Oximetry level < 94%	339	84.8
Respiratory rate >24 cycle/min	120	30
Conduction abnormalities:	110	27.55
Left anterior fascicular block	47	11.75
Right branch block	40	10
AF	15	3.7
Sino atrial block	9	2.1
Elongated corrected QT interval	97	24.2
Abnormal pulmonary auscultation sounds	62	15.5
Crackles	51	12.8
Ronchi	11	2.8
Sinus tachycardia	52	13
High temperature	15	4.1
Signs of cardiac failure	10	2.5

Radiological findings

Chest computed tomography (CT) was performed in 277 cases (69.3%).The mean delay between symptoms and chest computed tomography was 10 days [0-37days]. The most common findings were ground glass opacities (93.86%), consolidations (73.3%), followed by crazy paving (36%) (Table 2).

Table 2. Computed tomography results in older patients with Covid-19 pneumonia.

CT findings	N	%
Ground glass opacities:	260	93.86
Diffuse	215	82.69
Nodular	26	10
N/A	5	1.92
Consolidation	173	73.3
Crazy paving	144	36
Lymphadenopathy	46	16.6
Emphysema	43	15.52
Organizing pneumonia	40	14.44
Pleural effusion	36	12.99
Pericardial effusion	14	5.05
Centro-lobular micronodules	8	2.88
Bronchiectasis and bronchiolectasis	7	2.52
Architectural distortion	3	1.08
Pulmonary embolism	36	12.9
Distal	26	72.22
Bilateral	18	50
Spontaneouspneumomediastinum	1	0.36
Pneumothorax	1	0.36

Radiological lesions observed on CT scan were bilateral in 244 cases (88.08%) and had a basal predominance in 145 cases (52.34%). The parenchymal lung involvement was important in 116 cases (41.87%), severe in 66 cases (23.82 %) and critical in only 5 cases (1.8%).

Biological findings

Hyperleukocytosis was found in 127 cases (31.8%) while 25 (6.3%) patients had leucopenia. Lymphopenia and eosinopenia were found in 61.5% (N=246) and 47.8% (N=201), respectively. Anemia was noted in 34% of cases (N=136); 60.29% were females. It was microcytic in 79 cases (56.83%), normocytic in 52 (38.23%) and macrocytic in five cases (3.67%).Thrombopenia was found in 4.5% of cases (N=18). D-dimer adjusted to age were elevated in 274 patients (68.8%). Twenty-three patients (5.75%) had C-reactive protein (CRP) level within normal range (<10 mg/l). Procalcitonin level was positive in 32.4% (n=46) of cases.

Many disorders have been revealed including: liver analysis abnormalities (60.6%), phosphor and calcium metabolism disorders (54.6%), ionic disorders (43.8%), endocrinopathies (31.7%) and acute renal failure (25%) (Table 3).

Table 3. Metabolic disorders in elderly patients with Covid-19 pneumonia.

Metabolic disorder	N	%
Hepatic disorders	242	60.6%
Elevated transaminases	91	22.8
Cholestasis	87	21.8
Prolonged prothrombin time	64	16
Phosphocalcic disorders	218	54.6%
Hypophosphoremia	102	25.5
Hypocalcemia	93	23.3
Hyperphosphoremia	14	3.5
Hypercalcemia	9	2.3
Ionic disorders	175	43.8%
Hypokalemia	66	16.5
Hyponatremia	61	15.3
Hyperkalemia	40	10
Hypernatremia	8	2
Endocrine disorders	117	31.7%
Hyperthyroidism	71	18.6
Hyperparathyroidism	39	11,3
Hypothyroidism	7	1.8
Acute renal failure	100	25%
Myolysis	50	12.5%

Noteworthy, the most common endocrinopathy was hyperthyroidism in 71 cases (18.6%), 18 patients were retested three months after the infection and had normal results.

Immunological disorders

Direct Coombs test was investigated in 316 cases. It was positive in 36 cases (11.39%) without hemolytic anemia. Antinuclear antibodies were looked for in 175 patients and were positive in 40 cases (22.85%). Lupus anticoagulant (LA) was performed in 192 cases. It was positive in 82 patients (42.7%). Anti-B2Glycoprotein1 (anti-B2GP1) was performed in 202 cases and was positive in 15 (7.42%) of them. Anti-cardiolipin was positive in six cases (3%) out of 200 performed analysis. Double positivity was noted in four patients: three had anticardiolipin and lupusanti-coagulant positivity while one patient had both anti-B2GP1 and LA. No patient had threefold positivity of antiphospholipid antibodies.

Pneumonia severity assessment

Ratio of arterial oxygen partial pressure to fractional inspired oxygen (PaO₂/FiO₂ ratio)

PaO₂/FiO₂ ratio was calculated in 373 cases (93.3%). It ranged from 95 to 494.59 with a mean of 267.05.

QuickSOFA score (Q SOFA)

Two hundred and fifty patients (62.5%) had a q SOFA score of one, 123 patients (30.7%) had a q SOFA score of zero and in 27 cases (6.8%) the q SOFA score was two.

Oxygen requirements

Overall, 95% of patients (N=380) required oxygen during their hospital stay. Oxygen requirements at admission ranged from 1 L/min to 24 L/min with a mean of 3.27 L/min, 164 needed between one to two L/min, 111 needed three to four L/min, 37 needed five to six L/min and 71 needed more than six L/min.

Oxygen was administered via: nasal mask to 356 patients (89%), high concentration mask to 65 (16.3%), Opti flow to 13 patients (3.3%) and continuous positive airway pressure (CPAP) to 15 patients (3.8%).

Oxygen withdrawal delay ranged from one to 25 days with a mean of five days.

The duration of hospital stay ranged from one to 33 days with a mean of seven days.

Complications of COVID-19 pneumonia in older patients

Metabolic complications were the most frequent (N=148, 37%), followed by acute respiratory distress (N=80, 20%) and bacterial infection (N=62, 15.5%) (Table 3). Pulmonary embolism was observed in 36 cases (9%). Among patients with pulmonary embolism only two had concomitant deep vein thrombosis. Septic shock was noticed in 18 patients (4.5%). Cerebral stroke was noticed in 2 cases (0.5%).

Evolution

Full recovery was obtained for 334 patients (83.5%), 35 (8.8%) were referred to intensive care units, 28 (7%) died, two (0.5%) were discharged with home oxygen therapy and one patient (0.3%) was referred to another department. Of those who went to intensive care units (N=35), 12 died, eight recovered and were discharged, three returned to our ward and for the rest no information was available. Hence, the overall death rate was 10% (N=40).

Discussion

The COVID-19 outbreak affected people of all age categories. The older population, was strongly affected by the virus since the beginning, and age was considered an independent risk factor.⁴ Meanwhile, information about the characteristics of COVID-19 pneumonia and the potential prognostic factors in this specific group seem to be intriguing.

We designed a single large study including 400 elderly patients in order to report the clinical and radiological particularities of COVID-19 pneumonia in aged patients along with the rate of mortality and complications.

Since it is a retrospective study, determining risk factors and prognostic factors may not be accurate. Yet, we tried to extract factors associated with severe illness, fatality and mortality among our group of patients. Another limit of the study is that we did not present further information about the follow up of the discharged patients and the long-term impact of the hospitalization as perceived by this particular group of patients.

Many studies reported a majority of male patients being more prone to COVID-19 infection,¹⁷ which has been debated as susceptibility to infection is not gender dependent yet outcomes might be. That was explained by the higher levels of angiotensin-converting enzyme 2 (ACE2) in males and other behavioral factors including smoking and reaching out for medical help.¹⁸ In our study, we observed a slightly higher percentage of infected women which is in line with Lian et al. study of older COVID-19 patients.¹⁹ Among our patients, 30% had more than two comorbidities while 27% had two comorbidities, exceeding the rate that Wang et al. reported; 23.9% of their patients having two comorbidities or more.²⁰ In the latter study, hypertension (40.8%) was found to be the most common comorbidity followed by diabetes (16%) and our study revealed a higher prevalence of these two comorbidities 55.75% and 50%, respectively. An important fact is that these comorbidities were in many cases discovered during the hospital stay for COVID-19 pneumonia.

Fever is the most common symptom reported in the literature and it is a symptom leading patients to seeking medical help.¹⁵ In our study fever was reported by half of the patients yet it came in fourth place after dyspnoae, fatigue and cough. This disparity can be explained by the fact that fever may lack in frail elderly. The time lapse between onset of symptoms to admission was nine days, while the mean reported delay was six days in Córdova et al. systematic review.²¹ Peripheral blood oximetry levels below 94% were the most common abnormality in physical examination (84.8%) followed by dyspnoea (30%). Tachypnoea was found in 20.9% of the study population in Zhou et al. retrospective study in a single-center in Wuhan which was demonstrated to be an independent risk factor for mortality.²²

The most common ECG abnormalities were conduction disorders (27.55%) such as Left anterior fascicular block, right branch block, atrial fibrillation, sino-atrial block and prolongation of corrected QT interval (24.2%). AF were detected in 13% of COVID-19 patients admitted to the dedicated internal-medicine unit in Mele et al. study, that was explained by the older age of patients.²³ Long corrected QT interval have been mentioned in previous studies^24,25 even in absence of medication intake that induces QT interval prolongation such as azithromycin and hydroxychloroquine, which can be secondary to viral induced myocyte damage.

As in other literature reports,²⁶ the most frequent CT findings were ground glass opacities with mostly a peripheral and bilateral distribution. In elderly population, consolidations can be a predominant pattern compared to younger patients as well as atypical imaging findings.²⁷ The third most common pattern observed in the current study was crazy paving in 36% of cases yet its prevalence varies widely in the literature from 5% to 89%²⁸ and it has been associated to the inflammatory phase.²⁹ Pleural effusion has been mentioned in previous studies reaching up 23% in Okoye et al. study³⁰ which exceeded the percentage found in our study (12.99%) and was considered as a marker of disease severity and this higher proportion may be explained by the higher prevalence comorbidities like heart and renal failure. In our study, we noted many rare radiological findings like lymphadenopathy and pericardial effusion. Both have been associated to more severe pneumonia. Noteworthy, spontaneous pneumomediastinum was described in one case. It is a rare complication of Covid-19 pneumonia in the literature³¹ due to inflammation and alveolar injury with increased intra-alveolar pressure caused by coughing.

We found that 41.87% of patients had important involvement meaning that 26 to 50% of the lungs were damaged, while 23.82% had severe lesions and 21.29% had moderate involvement. Critical involvement was noted in only 1.8% of patients. While, six patients had no pulmonary lesions even though they had dypsnoae and hypoxemia. This may be explained by early imaging in the course of the disease.

In Ruch et al. series,³² moderate involvement was the most common (39.2%).The critical involvement represented 2.3% of cases. In the latter study, lung involvement > 50% was associated to early death or intensive care unit (ICU) admission while in our study we found that important parenchymal involvement corresponding to [26-50%] of lung damage was significantly correlated to death.

Moreover, C-reactive protein was elevated in 94.25% of patients which is corroborated by several studies.¹⁷ Procalcitonin was positive in 32.4% of cases, due to bacterial infection complicating the viral pneumonia course. However, in literature there was no significant difference reported between procalcitonin levels in elderly compared to young COVID-19 patients.

Lymphopenia was found in 61.5% of cases, similar to Wang et al. series.²⁰ Thrombopenia was also observed in 4.5% of cases and it was more important in dead patients. Eosinopenia found in 47.8% of cases, has been described as a biological marker of SARS-cov-2 infection.³³ Besides, D-Dimer levels adjusted to age were used to accurately interpret elevation due to common positive D-Dimers levels in older patients.³⁴ Elevated D-dimer levels were common (68.8%) and they not only predicted pulmonary embolism but also reflected the intensity of the inflammatory process and predicted poor prognosis.³⁵

The mean lactate dehydrogenase was 387 U/L, similar to the result reported in Chen et al. study.³⁶ They have been reported to be elevated in older patients and to help recognize severe cases of COVID-19 pneumonia.³⁷

Antinuclear antibodies positivity has been commonly reported in literature, with significantly higher titers in patients with poor prognosis. Out of 175 investigated antinuclear antibodies, 22.85% were positive close to the results reported by Peker et al. (18%).³⁸ Positivity of antinuclear antibodies is observed in other viral infections that can trigger auto-immunity in predisposed individuals.³⁹ Lupus anticoagulant was positive in 35.41% of tested cases, which is the most common anti phospholipid antibody (aPL), in line with the systematic review of Taha et Samavati,⁴⁰ who reported a 46.8% positivity of LA and a positivity of at least one aPL in nearly half of the infected patients. Single positivity of LA has been observed during the acute phase of COVID-19 infection, yet with no clear evidence in causing thrombotic complications.⁴¹ In our study, among patients with pulmonary embolism 19 had positive LA yet when retested only one case remained positive. Such result emphasizes the fact that single positive lupus anticoagulant is transient and it needs to be retested especially in patients with thrombotic complications in order to determine the duration of anticoagulation therapy.

The most common complications were metabolic complications in 37% of cases such as phosphocalcic disorders: hypophosphoremia and hypocalcemia were the most frequent. Hypocalcemia is common in severe disease and even non severe disease forms⁴² while hypophosphatemia may be associated with increased mortality.⁴³

Likewise other viral infections, Covid-19 infection can induce subacute thyroiditis either by the direct viral infection or post viral inflammation or by triggering the activation of auto-immune thyroiditis.⁴⁴ Thyroid dysfunction is usually reversible after remission. This fact explains the thyroid disorders, mainly hyperthyroidism noted in 18.6% of cases in this study that returned to normal in 25.3% of them. Acute kidney failure was noted in 25% of our study population, like the rate observed in Neumann et al. systematic review.¹⁷ Elevated transaminase and cholestasis have been observed in our study in 22.8% and 21.8%, respectively, even in the absence of viral hepatitis B and C infections. Liver damage induced by Covid-19 is common and the mechanisms of the dysfunction are due to the presence of ACE2 receptors causing cholangiocyte dysfunction.⁴⁵ Myolysis induced by Covid-19 infection have been reported in the literature with an overall incidence between 0.2-2.2%. Muscle enzymes elevation is more common.⁴⁶ Acute respiratory distress syndrome: mild acute respiratory distress syndrome was observed in 134 (64.73%) out of 207 calculated PaO₂/FiO₂ ratios, higher than the ratios reported by Zhou et al.,²² while moderate acute respiratory distress (ARDS) was noted in 17 (8.21%) of them which is lower than the aforementioned study.

Secondary bacterial infection occurred in 15.5% of cases higher than 4.9% reported in Neumann et al. review. A hypothesis to this result is the higher rate of comorbidities in our population study, long diagnosis and admission delays added to auto-medication before admission. The incidence of pulmonary embolism in the current study was 12.99%. The incidence rate varied in several studies and in a recent systematic review by Gong et al. the incidence in non-intensive care patients was 17%.⁴⁷ The endothelial injury, systemic inflammation, reduced mobility and platelets activation play a major role in the prothrombotic state in infected patients. Nonetheless, aPL positivity was not associated to increased risk of thrombosis or embolism.⁴⁰

Age has been known to be an important risk factor of poor outcome and mortality in Covid-19 patients, it rated 10% in our study. In a recent systematic review, reported death rates varied from 0% to 64% due to major differences in epidemiological characteristics of the studied cohorts.⁴⁸ Studies including patients aged 65 and above had higher mortality rates than our study.^49–51 The relatively low mortality rate in our study can be explained by the fact that the study included patients admitted to a medical department and did not include intensive care units patients add to that a relatively rapid admission. Another limit impacting the death rate is the lack of information regarding 15 patients referred to ICU. Patients aged 80 and over are rarely reported in literature²¹ and account for high death rates.⁵² In the current study we proved that advanced age is indeed predictive for in-hospital mortality.

Some studies reported that men had higher mortality risk than women.^53,54 The majority of our patients had comorbidities with the two most common being high blood pressure followed by diabetes which is in line with the literature⁵ although, they were not associated to higher mortality. Respiratory rate is known as a risk factor for mortality²² which was not proven in our study while dyspnoea was correlated with death, such in Zhou et al. study.²²

In Ruch et al.’s study³² lung involvement > 50% was associated with early death or intensive care unit admission yet in our study parenchymal involvement corresponding to [26-50%] of lung damage were well correlated to death. Oxygen needs exceeding six liters/min on admission were highly correlated to mortality, reflecting the severity of pneumonia and the need of more assistance. Prolonged QT has been proposed as a prognoses predictor by Akhtar et al.,⁵⁵ however this was not observed in our study. As in previous reports^56,57 elevated CRP levels were correlated to death. CRP has been considered as a marker of the cytokine storm in Covid-19 patients.⁵⁶ The length of hospital stay in our study was predictive of mortality, since it can be source of complications and it may promote transmission to healthcare workers.⁵⁸

Since it is a retrospective study, determining risk factors and prognostic factors may not be accurate. Yet, we tried to extract factors associated with severe illness, fatality and mortality among our group of patients. Another limit of the study is that we did not present further information about the follow up of the discharged patients and the long-term impact of the hospitalization as perceived by this particular group of patients. We are aware of selection bias, since people with comorbidities are more prone to hospitalization yet in comparison with other studies the high rate of comorbidities in our study group is certainly an eye-opening fact to the importance of systematically screening chronic diseases during hospitalization for acute disease.

In conclusion, Covid-19 became a major health issue worldwide. Elderly patients represent a particular category due to the vulnerable pre-existing state of patients, making the outcomes more severe. Thus, they need in-hospital medical care with longer stays, more ICU admissions with heavy outcome on health care systems.