Differences in Prevalence of Haematological Abnormalities on Presentation to Hospital in COVID-19-Infected Adult and Paediatric Patients: A Retrospective Multicentre Descriptive Study Open Access

Following the emergence of the new coronavirus 2019 (2019-nCoV), the world has seen a spectrum of presentations of the virus with different clinical outcomes. The clinical spectrum of COVID-19 varies from asymptomatic or mild disease to respiratory failure that requires mechanical ventilation and to multiple organ dysfunction syndromes or failure leading to death. Cytokine storm and thromboembolism have been identified as the main pathological processes of SARS-CoV-2 [1-3]. The enormous cytokine release in COVID-19 patients is linked to decreased lymphocyte count. Lymphocytopenia was present in 83.2% of patients on admission in a report from China and was linked to more severe cases [4]. Coagulation abnormalities such as prothrombin time (PT) and activated partial thromboplastin time (aPTT) prolongation, fibrin degradation product increase with severe thrombocytopenia were also noted and were commonly seen in patients with worse outcomes. Moreover, elevated dimer was also a frequent finding and levels above (1 μg/mL) were a strong and independent risk factor for death [2]. A meta-analysis reported lymphopenia of 13.5–16%, high procalcitonin of 36%, high C-reactive protein (CRP) of 19%, and high lactate dehydrogenase (LDH) of 29% in the paediatric population which was decreased compared to the adult population.

It has been postulated that the likely difference between the severity of the two populations is linked to the decreased angiotensin-converting enzyme 2 receptors and immaturity of the immune system in the paediatric population [5, 6]. The prevalence of haematological manifestations in the Middle Eastern population is not well documented.

The primary objective was to study the prevalence of haematological manifestations in the full blood count of patients admitted to hospital with COVID-19 infection within first 24 h of hospital admission.

The secondary objective was to compare the difference in the prevalence of haematological manifestation of COVID-19 and their clinical significance in adult and paediatric patients.

In retrospective descriptive multicentre study, 1,000 patient’s records were randomly selected of COVID-19-infected patients with confirmed pharyngeal PCR swab admitted at three Dubai Health Authority facilities, Rashid hospital, Dubai hospital, and Latifah hospital infection from 15 March–30 May 2020. After exclusion of patients as per study criteria, the final population size was 835 patients. Of these patients, 731 patients were adults (age 19 and above) and 104 patients were of paediatric age group (0–18 years). This study was approved by Dubai Scientific Research Ethics Committee (DSREC).

Patients below the age of 18 years were classified under the paediatric population [7]. The normal values for blood parameters were obtained from phase II of the pathology Harmonisation Project and Hematology of Infancy and Childhood [8-11] and American college of physician normal adult laboratory value reference as per the hospital standard.

Inclusion criterion was patients of all ages who tested positive for COVID-19 infection on nasopharyngeal PCR swab. Exclusion criteria were patients with pre-existing haematological disorders (like haematological malignancies, hemoglobinopathy, or immune thrombocytopenia), patients on warfarin or pre-existing coagulopathy, pregnant patients, and those with missing investigations.

All data were collected from patient’s electronic medical records through EPIC (Salama system) of the hospital and were then tabulated on an Excel sheet. Data recorded included the patients’ age, gender, white blood count, haemoglobin, platelet, absolute lymphocyte, neutrophil, monocyte, basophil, and eosinophil counts along with CRP, dimer, LDH, coagulation profile on admission to the hospital. Two researchers independently reviewed data to confirm the accuracy of the collected data.

Data were analysed using the IBM SPSS version 27.0 (IBM Corp., Armonk, NY, USA) under the supervision of an experienced statistician. All the categorical data groups were recorded as frequencies and compared using the Pearson χ² test. p values less than 0.05 represented statistical significance, and all reported p values were two-sided.

The identity of the patient was kept confidential at all times. Since it was an observational study, there was effect of the study on the treatment of the patient.

A total of 835 patients were studied. 731 were studied in the adult group, and 104 patients were studied in the paediatric group. Patient characteristics and final outcomes are reported in Table 1.

The mean ages were 44.7 ± 12.8 years and 4.9 ± 4.3 years in adult and paediatric groups, respectively, as described in Table 1. In the paediatric group, most patients 49% (50/104) were above the age of 3 years, while infants (below the age of 1 year) formed 22.1% (23/104) and toddlers (age 1–3 years) formed 28.8% (30/104) of the population.

There were no mortalities in the paediatric population, while 12.7% (93/731) of COVID-19 infections in adults resulted in death and 8.4% (62/731) of adult patients were transferred to another facility so their outcome is not known. The prevalence of at least one abnormal haematological parameter was 95.1% (794/835) on the first presentation to the hospital. After adjusting of age and gender, the prevalence of any white cell abnormality was 34.7% (290/835) (5.7% leukopenia, 9.6% leucocytosis, 25.4% lymphopenia, 5.5% neutropenia, 16.4% neutrophilia, 7.3% monocytosis, and 1.2% eosinopenia). A prevalence of 15.3% (128/835) anaemia, 9.5% (79/835) thrombocytopenia, and 4.3% (36/835) thrombocytosis was also observed. The prevalence of other abnormal blood parameters was C-reactive protein 69.5% (573/835), D dimer 57.5% (280/835), high LDH 52% (383/835), high ferritin 72.1% (452/835), high international normalized ratio 5.1% (38/835), prolonged PT 32.2%(240/835), and prolonged aPTT 35.6% (264/835). Most of these findings were more common in adults as described in Table 2.

A significant difference in the prevalence of these abnormalities was evident between the two populations as described in Table 2. The haematological abnormalities were much more prevalent in adults. Still, interestingly paediatric population had higher incidence of neutropenia (4.5% in adults vs. 12.5% in paediatrics), eosinophilia (0.4% in adults vs. 3.8% in paediatrics), and monocytosis (6.4% in adults vs. 13.5% in paediatrics) (p < 0.001).

Multiple prior studies have established the effect of the SARS-CoV-2 on the hematopoietic system [2]. The key goal of our study was to explore some of these haematological manifestations associated with COVID-19 infection and their prevalence in our population. To the best of our knowledge, this is one of the largest and only observational studies of this region which described the haematological abnormalities at presentation while comparing these findings in adult and paediatric populations.

Adult patients formed the majority 87.5% of the patients in the study. In the adult population, most patients were males 87.4%, while 52.8% of patients were male in the paediatric population. The paediatric population had no significant comorbidities, while diabetes mellitus and hypertension were the most common comorbidities in the adult population at 27.3% and 21.5%, respectively.

In the paediatric group, most patients 49% were above the age of 3 years, while infants (below the age of 1 year) and toddlers (age 1–3 years) formed 22.1% and 28.8% of the population, respectively. All the patients in the paediatrics group had a mild infection without any mortality. Perhaps, the presence of comorbidities also contributed to mortality in adult patients.

The most common blood abnormalities at presentation were noted to be increased inflammatory makers including high ferritin (72.1%), high CRP (69.5%), high LDH (52%), followed by deranged coagulation, high dimer (57.5.%), and prolonged PT (32.2%) and (aPTT 35.6%); other common findings were lymphopenia (25.4%), neutrophilia (16.4%), anaemia (15.3%), leucocytosis (9.6%), monocytosis (7.3%). It has been proposed that COVID-19 infection causes necrosis and apoptosis of the lymphocytes due to the inflammatory cytokines. Lymphopenia has been associated with severe infection and increased mortality [12, 13].

As reported in a meta-analysis, high CRP (58.3%, 95% CI: 21.8–94.7%), high LDH (57.0%, 95% CI: 38.0–76.0), lymphopenia (43.1%, 95% CI: 18.9–67.3), and high erythrocyte sedimentation rate (41.8%, 95% CI: 0.0–92.8) were the most prevalent laboratory results. These results are comparable to our findings [12].

Cytokine storm and hyperinflammation triggered by the COVID-19 underly the pathophysiology of severe COVID-19 infection. There is evidence that monocytes and macrophages play a role in fibrotic changes in the lungs of critically ill patients. It has also been proposed that a condition similar to macrophage activation syndrome is present in patients with severe COVID-19 infection. The patient’s innate immune response has a critical role in the outcome of the infection. Perhaps, the immature immune system of the children explains the milder severity of disease in this population apart from the decreased angiotensin-converting enzyme 2 receptors compared to the adult population [13-15].

We noted a significant difference in the prevalence of these findings in adults and the paediatric population (Table 2). These abnormalities were much more common in adults compared to paediatric patients, but interestingly paediatric population tended to have a higher incidence of neutropenia (p < 0.000). Moreover, similarly, leucocytosis was observed in 10.9% of the adult population, while leukopenia was more evident in the paediatric population (6.7%) (p 0.002). The neutropenia and leukopenia noted in our study may be a reflection of congenital neutropenia syndromes which is prevalent in our population.

Lymphopenia was observed in 28.5% of the adult population and only in 3.8% of the paediatric group. The prevalence of lymphopenia was less in our studied group compared to the previously reported 63% and 83.2% in a study reported from different parts of China [4, 16]. Furthermore, comparison of these findings with data in the Middle East could not be done as data on this topic in this population were limited.

It has been observed that higher white blood cell levels, neutrophilia, lymphopenia, and lower monocyte counts could be because of superimposed bacterial infection and also because of the effect of different cytokine releases during the infection, and these findings were more prevalent in the adult population which had 12% (93/741) deaths compared to no deaths observed in the paediatric population [15, 17]. These findings of higher lymphocyte count, lower neutrophil count, and lower inflammatory markers in paediatric patients were noted in other studies. Neutrophilia has also been noted in paediatric patients with MIS-C, but we did not have any patients who developed MIS-C in our study [17].

Many studies have linked haematological findings like lymphopenia with complications and poor outcomes of COVID-19 infections. These findings can be used as markers to identify patients who might benefit from early treatment and close follow-up especially as the cases of COVID-19 rise again [16, 18].

This study has several limitations as it is a retrospective study; many blood tests which were not ordered in the past like interleukin 6 which is linked to severe outcomes could not be studied. None of the paediatric patients had a severe COVID-19 infection, so no correlation between the laboratories and severity of infection could be studied.

Like most viral infections, SARS-CoV-2 infection affects the immune system of the body and generates an inflammatory response which can be detected on simple routine blood tests. Many of these findings like lymphopenia, neutrophilia, and high inflammatory markers have been linked to severe COVID-19 infections and worse outcomes for the patients. These abnormalities were prevalent in the adult population, and some findings like neutropenia were more common in the paediatric population likely reflecting congenital/benign neutrophil disorders in the population and different immune responses leading to lesser inflammation and damage. We recommend further studies to compare the difference and clinical significance in haematological manifestations with different COVID-19 variant strains to identify patients prone to worse outcomes who might benefit from earlier treatment.

The authors would like to thank and acknowledge all valuable support from Zoubia Mohamed, Sally Awad Mohamed Rahma, Harshitha Janardhan, Lana G. Adra, Maymona Motasim Mohammad, Rommana Mehdi, Anam Ahsan, Diary Mohammad, and Marwan Zidan.

The research complies with the guidelines for human studies and was conducted ethically in accordance with the World Medical Association Declaration of Helsinki. Ethical approval was taken from Dubai Scientific Research Ethics Committee (DSREC) Number DSREC/RRP/2020/22, dated August 30, 2020. Consent was taken from patients as a part of general informed consent signed on admission to Dubai Health Authority which states de-identified data can be used for research and the study was conducted maintaining full patient confidentiality.

The authors have no conflicts of interest to declare. The authors report no relationships that could be construed as a conflict of interest.

No funding was received for this study.

Saubia Fathima and Laila Al Dabal are responsible for conception and design of the study; data analysis and interpretation were done by Saubia Fathima and Salma Rahma. Saubia Fathima, Farah Chughthai, and Salma Rahma drafted and revised the final article. All authors read and approved the final draft of the manuscript.

The data that support the findings of this study are not publicly available as data were collected from patients’ electronic medical record and can compromise the privacy of research participants but are available from the corresponding author upon reasonable request.