Introduction

The COVID-19 epidemic began in late December of 2019 in China. The disease spread rapidly to other parts of the world and was declared a pandemic by the World Health Organization on March 11, 2020. The first case in Poland was reported on March 4, 2020 and 10 days later, the first infected person in Pomerania Province was identified. The first case of SARS-CoV-2 infection in a hemodialysis patient in Pomerania Province was diagnosed on April 7, 2020.1 Dialysis patients have a high rate of comorbidities and a high frailty index; they are often elderly and have lowered immunity, which puts them at risk of SARS-CoV-2 infection and a severe course of the disease.2 In addition, rapid spread of the virus in this group of patients may be facilitated by the fact that hemodialysis units, waiting rooms, and means of transportation are collective spaces with a high traffic of people. Although specific guidance on the management of patients requiring dialysis during the COVID-19 pandemic is available, there is still a limited number of published long-term studies in well-defined hemodialysis populations.3,4 A preliminary report from a single region of Poland showed a high mortality rate among hemodialysis patients with COVID-19 during the first wave of the pandemic in the spring of 2020.5 Early case series reports from different countries suggested that case fatality in hemodialysis patients varied from 20.3% to 29.6%, with an incidence rate higher than in the general population.6-10 Most of the published reports were single-center studies of short duration. To shed more light on this issue, we performed a retrospective multicenter cohort study including all hemodialysis patients in Pomerania Province to gather key information on the disease trajectory and patients’ outcomes during the first and second waves of the COVID-19 pandemic in Poland, until the commencement of the national SARS-CoV-2 vaccination program.

Patients and methods

Pomerania Province is located in the northern part of Poland (Figure 1). On December 31, 2019, its population amounted to 2 343 928 inhabitants, including 1 881 844 adults. A total of 1319 patients were on chronic dialysis at this time, with about 95.3% receiving hemodialysis and 4.7% receiving peritoneal dialysis. Chronically hemodialyzed adult patients from Pomerania Province are treated in 15 dialysis units, of which 4 are public and 11 private (Figure 1).11 They are all financed by the National Health Fund. This was a retrospective multicenter cohort study including all adult patients receiving chronic in-center hemodialysis in Pomerania Province on December 31, 2019 and all new patients who were starting long-term hemodialysis between January 1, 2020 and January 31, 2021. The control group was the adult population of Pomerania Province as of December 31, 2019. In 2020, there were no significant changes in its size. Changes in its structure were slow enough not to have significantly affected our results and conclusions. Each dialysis center was asked to report all new cases of COVID-19 and all deaths due to the disease since the beginning of the pandemic until January 31, 2021. A confirmed case was defined as a positive result of a SARS-CoV-2 reverse transcriptase–polymerase chain reaction test of nasopharyngeal / oropharyngeal swab specimens. The assessment of whether or not death was due to COVID-19 was performed by healthcare personnel at each dialysis center. In principle, testing in the majority of units was prompted by suggestive clinical signs. Single dialysis units periodically tested their entire dialysis population. Primary outcome of the study was the incidence of SARS-CoV-2 infection and death due to COVID-19 in adult hemodialysis patients. The results were compared with the incidence as well as mortality and fatality rates observed in the same time frame in the general adult population of Pomerania Province. Data on the number of daily new cases and COVID-19 mortality in the general population were taken from the reports of administrative and health authorities (Voivodeship Office, Ministry of Health, Local Sanitary Inspectorates, and Central Statistical Office) and responses to requests for access to public information. During the pandemic, the dialysis unit at the 7th Navy Hospital in Gdansk, as a result of the decision of the health authorities, was repurposed into a designated hemodialysis center adapted to take care of patients with COVID-19, and patients who had been receiving chronic hemodialysis there before the pandemic were transferred to other nearby centers for treatment. Therefore, for the purpose of the present analysis, this dialysis unit was not taken into account and individuals who had been chronically hemodialyzed in this center on December 31, 2019 were considered as patients of the units to which they were transferred in 2020.

Figure 1. Distribution of the 14 dialysis units in Pomerania Province participating in the present study and the size of the hemodialysis and general adult populations of the region over the study period

Ethical approval for the study in all dialysis units was obtained at Medical University of Gdansk (NKBBN/2014/2021). The study is part of the “COVID-19 in Nephrology” (COViNEPH) project focusing on the nephrological aspects of COVID-19, in particular the epidemiology, prevention, disease course, and treatment.

Statistical analysis

Descriptive statistics were used to outline the clinical characteristics of patients. Continuous variables were presented as mean (SD) and they were compared with the t test. Categorical variables were presented as relative and absolute frequencies. The χ2 test was used to compare the categorical variables where appropriate. A 2-tailed P value of less than 0.05 was considered significant. The age distribution of hemodialysis patients was substantially different from that of the general population; hence, standardization for age was necessary to compare the incidence, mortality, and fatality rates. The age-standardized mortality / fatality / incidence ratio was then calculated as the ratio of observed to expected number of deaths / cases, as appropriate. The general adult population of Pomerania Province was selected as the reference. No more than 0.75% of data on COVID-19 cases in the general population were missing or incomplete. Because the total number of cases was large, those with missing data were removed from the analysis. Data were analyzed using the STATISTICA software package, version 12.0 (Stat Soft Inc, Dell Software, Tulsa, Oklahoma, United States).

Results

Data were provided by all 14 dialysis units. The total number of exposed patient receiving chronic hemodialysis in the study period was 1567. Of these, 352 (199 men, 153 women) had confirmed COVID-19 and 262 (74.4%) were hospitalized. Overall, 107 patients died, including 49 women and 58 men. The absolute cumulative incidence of COVID-19 in the study period was 22.4% among hemodialysis patients and 5.17% in the general adult population (P <⁠0.001) (Table 1). There was no difference in the incidence between men and women receiving hemodialysis (199 [21.1%] out of 941 vs 153 [24.4%] out of 628, respectively), but the incidence varied widely by dialysis units, ranging from 9.3% to 36.9%. After standardization for age, cumulative incidence was 3.98-fold higher in hemodialysis patients than in controls. Peaks in the trends of new cases coincided in both groups. The increase in the number of new cases during the autumn wave of the pandemic was more marked in the hemodialysis population compared with controls. The decline of the number of new cases in hemodialysis patients was sharper and occurred earlier than in the general population (Figure 2).

Table 1. Number of COVID-19 infections, incidence, mortality, and fatality rates in hemodialysis patients and the general adult population of Pomerania Province in the period from March 14, 2020 until January 31, 2021

Variable

Hemodialysis population

General adult population

Population, n

1569

1 881 844

COVID-19 cases, n

352

97 148

Age of infected patients, y, mean (SD)

67.99 (13.09)

48.57 (17.37)a

COVID-19 deaths, n

107

2212

Age of deceased patients, y, mean (SD)

71.16 (12.91)

76.12 (11.68)a

COVID-19 incidence rateb, %

Crude

22.43

5.17a

Age-adjusted

20.45

–

COVID-19 mortality rate per 1000c

Crude

68.2

1.17a

Age-adjusted

12.85

–

COVID-19 fatality rated, %

Crude

30.4

2.27a

Age-adjusted

11.69

–

a P <⁠0.001

b Incidence rate was calculated as follows: (number of new cases / number of exposed people) × 100.

c Mortality rate per 1000 was calculated as follows: (number of COVID-19 deaths / number of exposed people) × 1000.

d Fatality rate was calculated as follows: (number of confirmed COVID-19 deaths / number of COVID-19 cases in exposed population) × 100.

Figure 2. Number of new cases of SARS-CoV-2 infection among hemodialysis patients and in the general adult population of Pomerania Province per reporting week

The absolute cumulative mortality rate due to COVID-19 per 1000 hemodialysis patients was 68.2 as compared to 1.17 per 1000 adults in the general population (P <⁠0.001). After adjusting for the significant confounding effects of age, the COVID-19 mortality among hemodialysis patients was 10.98-fold higher than in controls (Table 1).

Fatality rates in hemodialysis patients infected with SARS-CoV-2 and controls were 30.4% and 2.27%, respectively (P <⁠0.001). After standardization for age, the fatality rate in the former group remained 5.15-fold higher than in the general population. Older age significantly affected the fatality rate in both groups; it reached 43.81% in hemodialysis patients aged 75 years or older (P = 0.003) (Table 2). Among men and women receiving hemodialysis, the fatality rates were 29.14% and 32.03%, respectively, compared with the respective rates of 2.96% and 1.71% among men and women in the general population (P <⁠0.001).

Table 2. Absolute fatality rate of COVID-19 in different age groups

Fatality ratea, %

Age group

18–44 y

45–64 y

65–74 y

>74 y

General adult populationb

0.09

0.84

6.08

10.79

Hemodialysis populationc

18.52

21.33

27.59

43.81

a Denominators in these calculations were cases of SARS-CoV-2 infection among the relevant analyzed subpopulations.

b Pearson χ2 test, P <⁠0.001

c Pearson χ2 test, P = 0.003

Discussion

Patients requiring chronic in-center hemodialysis are a specific and particularly vulnerable population during the pandemic. The necessity to receive treatment at the dialysis center 3 times a week precludes them from staying isolated in their homes. They have to routinely interact with drivers, other patients, and members of the healthcare team; therefore, they are identified as individuals at high risk of COVID-19. Dialysis units create the perfect environment for the rapid spread of SARS-CoV-2 infection. Moreover, patients undergoing dialysis have high rates of comorbidities, they are often older and have varying degrees of immunosuppression, which puts them at risk of developing very severe forms of the disease.2,3

The incidence rate of COVID-19 was very high among the analyzed population of hemodialysis patients. The absolute cumulative incidence was 22.4% and varied widely by dialysis units. In the period from the beginning of the pandemic until the commencement of the national COVID-19 vaccination program, it turned out to be almost 4-fold higher than in the general population of Pomerania Province. Of note, however, this difference may be somewhat overestimated due to the SARS-CoV-2 screening model adopted in Poland, which was focused only on symptomatic patients, thus potentially affecting the reporting of new cases, especially in the general population. On the contrary, higher infection rates among hemodialysis patients may be due to easier access to SARS-CoV-2 testing. Moreover, some dialysis units performed periodic screening tests among all hemodialysis patients or those who had had contact with persons with confirmed SARS-CoV-2 infection.

Since the beginning of the epidemic, various actions to limit the spread of the virus have been implemented by dialysis centers3 and the incidence of COVID-19 in the units ranged from 9% to 37%. These large variations may be due to the too liberal policy adopted by some units with regard to safety measures, noncompliance of some patients or single cases of careless behavior. On the other hand, some findings on the incidence trajectory may indicate the effectiveness of the containment measures in the hemodialysis population assessed as a whole. The first wave of the COVID-19 pandemic in the spring of 2020 was flat and was probably held back by early introduction of national lockdown. The increase in the number of new cases during the second wave of the epidemic in autumn occurred earlier in the hemodialysis population as compared with controls; a similar phenomenon was observed in the dialyzed population of Ontario.8 Interestingly, at the end of the study period, an increase in the number of cases, indicating the beginning of a third wave, was noticeable among hemodialysis patients, but not yet visible in the general population (Figure 2). Presumably, the density of the dialysis population can be associated with higher rates of SARS-CoV-2 transmission and the role of density is particularly vital in the initial phase of the epidemic.12 The number of new cases increased considerably between mid-September, when the beginning of the post-vacation wave of infections was observed, and mid-November, when the second wave of infections peaked in the 2 groups at the same time. The decline in the incidence curve in the hemodialysis population started earlier and was faster than in the general population, similarly to the data from Ontario, United States and Flanders region, Belgium.8,9 Earlier suppression of the disease dynamics in hemodialysis patients may have been influenced by prevention plans implemented in all dialysis units, including, inter alia, the universal use of surgical masks by patients during transport and dialysis, quarantine after coming into contact with an infected person, assessment of patients by the staff by means of an epidemiological interview and body temperature measurement before admission to the unit, early examination of any patient suspected of COVID-19 (nasopharyngeal swab), and transfer of patients with COVID-19 to the designated unit.

Data from the annual report of the Polish Nephrology Registry indicated a large increase in the overall mortality among hemodialysis patients in Pomerania Province in 2020 as compared with 2019 (reaching 25%), with COVID-19 being responsible for over 30% of all deaths (unpublished data).11 The results of our study seem to confirm that the COVID-19 pandemic is largely responsible for the high mortality rate in this population. The fatality rate in our cohort of hemodialysis patients with COVID-19 was found to be 30.4%, ranging from 18.5% in the youngest and middle-aged patients to over 43% in the oldest individuals (≥75 years old). Old age is associated with a proinflammatory and procoagulant condition termed “inflammaging,” which provides an explanation for the strong correlation between age and COVID-19 mortality both in the general population and in hemodialysis patients.8,9,13 Unlike other studies,9 we did not find a clear relationship between male sex and mortality in hemodialysis patients; it was, however, noticeable in the general population of Pomerania Province. Lack of such a correlation in patients on long-term hemodialysis was also reported in studies from the United Kingdom and Ontario, United States.6,8

As mentioned earlier, the substantially greater fatality rate than that reported for the general Pomeranian population was probably caused by a higher frequency and severity of comorbidities, greater frailty, less efficient immunological response, and, primarily, the higher mean age of hemodialysis patients, considered the strongest predictor of poor prognosis and mortality.14,15 Due to the high incidence and severe course of the disease among patients requiring hemodialysis, the mortality rate in this group was almost 11-fold higher than in the general population after adjustment for the confounding effects of age. An early report from Wuhan indicated a milder disease in hemodialysis patients.16 However, more recent reports are suggesting worse outcomes, which is concurrent with our results. The Brescia region of Italy was early and severely affected by COVID-19. An observational study from 4 dialysis centers in this region of Italy showed a fatality rate of 29%.17 Multicenter observational reports from Brazil, New York, Ontario, Scotland, United Kingdom, France, and Flanders region of Belgium reported fatality rates ranging from 21% to 29.6%.7-9,13,18,19 Preliminary results from the ERACODA database, designed to prospectively collect detailed individual data of European dialysis patients with COVID-19, showed that COVID-19–related 28-day case-fatality rate was 25% for all patients and 33.5% for those requiring hospitalization.20

The main strength of the present study is the completeness of data. All the dialysis units responded to the request and reported all COVID-19 cases in their units on a daily basis. Our study covers the entire period of the pandemic in Pomerania Province caused by the original strain of the virus. Variant B.1.1.7, which appeared in February 2021 and became dominant in March 2021 in Pomerania, has a higher transmission rate and will probably significantly alter the epidemiological indicators. To our best knowledge, this is an analysis of a hemodialysis population with the longest study period to date, covering the time span since the beginning of the pandemic throughout 2 waves of SARS-CoV-2 infections until the commencement of the vaccination program, which took place in our region at the end of January 2021. Daily reporting of cases allowed for accurate representation of incidence trajectory over time. Standardization for age rendered the results on the mortality burden of COVID-19 among hemodialysis patients more reliable.

However, our results should be interpreted with caution, bearing in mind the following limitations. First, the nonsystematic SARS-CoV-2 screening model adopted in Poland favors the detection of the more severe cases only; therefore, the presented results may have underestimated the incidence rate and overestimated the fatality rate. Moreover, a proportion of people in the general population with mild symptoms of COVID-19 often chose not to undergo testing to avoid quarantine, thus further intensifying this effect. Second, as mentioned above, easier access to SARS-CoV-2 testing together with periodic testing of all hemodialysis patients performed in some units could have overestimated some results, particularly when compared with the general population. It should be kept in mind that numerous cases, also in the hemodialysis population, are asymptomatic. In a study by Anand et al,21 only 9.2% of dialysis patients with SARS-CoV-2 antibodies were reported as previously confirmed by a polymerase chain reaction test,21 and screening of all patients by means of a SARS-CoV-2 polymerase chain reaction test in a Spanish dialysis center revealed that 39% of individuals with positive test results were asymptomatic.22 An important limitation is lack of data on the prevalence of risk factors for mortality among the study patients, for example, hypertension and diabetes as well as dialysis vintage and the medications used. However, it can be assumed that the majority of patients had hypertension and diabetes, since the data from the Polish Renal Replacement Therapy Registry show that diabetes and hypertension are the main causes of end-stage kidney disease in 14% and 28% of cases, respectively.11 The lack of these data resulted from the survey design of the study. In addition, the study was primarily aimed at assessing incidence and mortality rates, and not the factors affecting these parameters. To ensure the completeness of the daily incidence and mortality data collected from numerous dialysis units, which was accomplished, we have abandoned the requirement of reporting detailed data on patients’ characteristics by the centers.

Despite some limitations, the results of this study show the extremely high mortality rate among COVID-19 patients on in-center hemodialysis and they are important with regard to supporting the collective efforts to minimize the risk of virus transmission in this very vulnerable patient group.