Our findings demonstrate the effects of socioeconomic, ethnic and cultural determinants on the epidemiology of COVID-19 in Israel.
Our analysis shows that more tests per population were performed in the most deprived populations compared with the least deprived ones. This might reflect the early awareness of the MOH to communities with high prevalence of confirmed cases, particularly the Ultra-Orthodox communities. As a result, epidemiological investigations and testing were directed to those communities.
A striking finding is the increasing rate of confirmed cases with decreasing SE categories, with rates more than double in the lowest category compared with the higher ones. A similar social gradient in health has been documented in New York City (NYC), where the likelihood that a test was positive was greater in poor neighborhoods, as well as in neighborhoods with larger households or a predominantly black population.17
CFRs were lower in low SE categories compared with higher SE categories. Due to the small number of COVID-19-related deaths in Israel, mortality data should be interpreted with caution. CFR in the Israeli population that was analyzed here (7·7 Million) was 1·37%, compared with 4·2%, 5·8% and 13·7% in Germany, US and Italy, respectively.18 Possible explanations for the relatively low mortality rate observed in Israel include the country's good healthcare system with universal coverage of care,19 its younger population compared with other OECD members,20 which may be less affected by COVID-19,21 and early establishment of national measures. These measures included stopping of incoming flights and required 14-day isolation of those who arrived from abroad, social distancing measures, with an emphasis self-isolation of the elderly, a national state of emergency and almost complete lockdown between March 19th and May 3rd, 2020. In addition, frequent military operations and emergency situations have prepared the Israeli healthcare system for crises, contributing to its ability to effectively care for severely ill patients.
As reported for other countries,22 it is possible that the number of COVID-19-related deaths in Israel have been under-estimated due to deaths that occurred at home prior to a COVID-19 diagnosis. During the first weeks of the pandemic, all health maintenance organizations in Israel advised their members to shift to telehealth platforms as the main venue for communicating with their health providers, instead of visiting their clinics. Since the most deprived patients are less comfortable with digital infrastructures,23 as well as less health literate, they might be more vulnerable to disease deterioration. However, as all Israeli citizens have access to public healthcare,19 the possibility of major under-reporting of infection and death is unlikely. Moreover, during the study period, all-cause mortality, reported by the Israel Center for Disease Control, was within range of the multiannual average,24 supporting this assumption.
Experience gained in prior health crises as well as from emerging reports from the COVID-19 pandemic, showed that ethnic minorities and depraved populations are more susceptible to being affected by the disease.4, 5 Therefore it was expected that the Arab ethnic minority in Israel, which is demographically characterized by being less affluent, having larger families, greater population density, residence in geographical peripheries, and a higher prevalence of co-morbidities, would demonstrate higher infection rates and more severe disease course compared with the general population. However, only moderate differences in testing were noted between the Jewish and Arab populations (5·6% vs. 4·1%), which might be explained by less access to healthcare services in a population of which more than half reside in geographical peripheries9 and fear of stigma and social, cultural and economic consequences of being found "positive". To help overcome these obstacles, political representatives urged the MOH to increase testing among the Arab population.
The strikingly low infection rates among the Arab population – almost five times lower compared with the Jewish population – may be attributed to the following: a) Obedience of the population, which is predominantly (85%) Muslim9 and religious, to the official religious instructions (named "Fatwa" in Arabic) conveyed by the religious leaders ("Mufti"). The religious leaders instructed the population to follow the instructions of the MOH and relevant official bodies; to postpone or shorten cultural events; to perform funeral services with minimal participation and to avoid any kind of gathering together.25 They also closed all of the mosques, instructing people to pray at home instead, especially during Ramadan, which began on April 23rd. b) Arab men and women are employed at high rates in the Israeli healthcare system as physicians, nurses and pharmacists - the most prominent occupations of the COVID-19 crisis – contributing to the feeling that the Israeli Arabs citizens are full partners in this struggle.9, 26 This high rate of knowledgeable people could possibly contribute to maintaining social distancing and other measures to fight the disease. c) Traditionally, Arab families take care of the elderly at home and do not place them in residential care facilities, reflecting a culture that is more tolerant toward older adults.27, 28 This may have been a protective factor, as a third of all COVID-19 deaths in Israel during the analysis period were documented among elderly living in residential homes.29 d) Finally, the political alliance of the main Arab-majority political parties in Israel – the Joint List – showed involvement by focusing on issues relating to all Israeli vulnerable sectors.26 Their voice was loudly heard among their voters and among policy makers.
Despite what seems to be an optimistic situation among the Israeli Arabs, one must consider the possibilities of clusters of undiagnosed infections, particularly in remote, segregated communities that are isolated geographically. Lower access to healthcare services might also explain late testing, and consequently a relatively late explosion of cases. The last week of April and the first days of May, saw a surge of confirmed cases in some Arab localities, such as the Bedouin locality of Hura (Figure 2); however, it is too early to judge if the Moslem Arabs followed their leaders' instructions to avoid large gatherings during the month-long Ramadan holiday. Such gatherings may reverse the seemingly lower prevalence of COVID-19 demonstrated in the Arab society during the first month of widespread disease in Israel.
The presentation of COVID-19 in the Ultra-Orthodox community, which bears socioeconomic similarities to the Arab community, was completely different, manifesting early with very high infection rates, as shown in the two localities that are predominantly Ultra-Orthodox – Bene Beraq and El'ad (Figure 2). The community participates only minimally in secular affairs and opposes features of public Israeli life, such as the national government and modern law. Most Ultra-Orthodox people perceive an external threat from the surrounding secular society and are in constant need to react to and defend themselves from such threats. This translates into segregation from the rest of Israeli society.30 The emergency orders and restrictions issued by the government hit the most fundamental components of their identity – group studies of the bible (Torah), prayer in a quorum of ten men and ritual baths. These traditions were stronger than the biblical rule that everything should be put aside for the sake of saving lives.26 The instructions for strict social distancing were rejected because they were not endorsed by the community's own leaders. Moreover - the senior leader of this community loudly opposed the instructions to close schools and told his followers to continue group learning. These Ultra-Orthodox leaders stopped opposing the government's instructions only after massive infection spread was seen in their communities. To curb the spread of infection and flatten the escalating curve, the Israeli government imposed hermetic quarantine on some Ultra-Orthodox neighborhoods.
Our analysis has several limitations: First, data were analyzed by locality, rather than by individuals. This ecologic approach misses possible differences in the composition and disease patterns within localities. Second, small towns and settlements, comprising less than 5000 residents, were not included in the MOH database and therefore were not analyzed. No data exists to evaluate this possible bias on disease measures. Third, data on localities comprising mixed Arab-Jewish ethnicities were excluded from the analysis, because it was not possible to distinguish between the ethnicity of individuals in the database. Arab communities in large mixed cities might present with different disease patterns. Last, after April 26th, 2020 mortality data and ethnicity were no longer reported at the localities level in the MOH database; however, due to intensive media coverage of fatalities, including the naming of most of those who died, we believe that large numbers of Arab deaths could have not escaped public awareness.
Our analysis shows that we should treat assumptions about ethnic minorities with caution as it is not always possible to predict how such communities would react in times of crisis. The Israeli Arab community mostly demonstrated responsible behavior, following governmental instruction that were mediated by the community's religious leaders. Although, the Ultra-Orthodox community did not abide by the government's instructions because it regards its religious leaders' instructions as more relevant, it is important not to stigmatize a community "en-bloc". Such labeling of the whole community is unjust and counterproductive. Instead, the more cooperative leaders of the community should be strengthened so that their voice may be heard.