History and current status of cancer registration in Russia

Background: Russia, then part of the Union of Soviet Socialist Republics (the USSR), introduced compulsory cancer registration in 1953, but a clear overall contemporary description of the cancer surveillance system in Russia is not available. Methods: We summarized historical landmarks and the development of the standards of classification and coding of neoplasms in Russia and described current population-based cancer registries’ (PBCR) procedures and practices. Results: Cancer registration is organized according to the administrative division of the Russian Federation. More than 600,000 cases are registered annually. All medical facilities, without exception, are required to notify the PBCR about newly diagnosed cases, and each regional PBCR is responsible for registering all cancers diagnosed in citizens residing in the region. The data collection can be described as passive and exhaustive. Hematological malignancies, brain, and CNS tumors are often not referred to cancer hospitals in some regions, explaining the problems in registering these cancers. Conclusion: Russia’s cancer registration system is population-based, and practices seem to be generally internationally comparable. However, coding practices and national guidelines are still outdated and not up to the most recent international recommendations. Further analyses are needed to assess the comparability, validity, completeness, and timeliness of Russia’s PBCRs data.


Background
Russia, then part of the Union of Soviet Socialist Republics (the USSR), introduced compulsory cancer registration in 1953. But it was not before 1963 when the USSR updated data collection and processing procedures to be comparable to those of Western cancer registries [1]. International Agency for Research on Cancer (IARC) Technical Report #35 summarized cancer registration practices in the Newly Independent States of the former Soviet Union (Belarus, the Russian Federation, and Ukraine) in 1998 [2], highlighting some important limitations. This report mentioned infrastructure for data collection, the introduction of international classifications for coding, and systemic monitoring of quality as prerequisites for obtaining data for research purposes.
Researchers and healthcare professionals can obtain cancer incidence and mortality statistics for Russia from several sources, but a clear overall description of the cancer surveillance system is not available. Systematic description of the cancer registration system is a critical step in understanding cancer statistics. For example, such a report has recently been published for the Nordic countries to explain similarities and differences despite high-quality cancer registries in all countries [3]. Here, we aimed to describe the history and current state of cancer registration procedures and practices in Russia.

History of cancer registration in modern Russia
We summarized historical landmarks and the development of the standards of classification and coding of neoplasms in Russia in Fig. 1. The Ministry of Health introduced definitions for Regional and National Cancer Registries in Russia only in 1996 and IARC described cancer registration procedures in Russia in 1998 [2,4]. Although the Ministry of Health's Order #420 introduced definitions of regional and national PBCRs in 1996, in practice, the national cancer registration system was re-established in Russia in 1999 when another Order #135 was published [5]. PBCRs were established independently in all the regions, and they are obliged to follow the common framework of national legislation. Order #135 provided instructions and the classification needed for cancer registries' appropriate functioning across the country. It also introduced the ICD-O-2 morphological classification, which most of the Russian PBCRs are still using.
Several regional PBCRs had already collected incidence data before the new cancer registration system's full-scale introduction in 1999. Saint-Petersburg formally established the first PBCR in Russia in 1993. It was the only Russian cancer registry in the IARC "Cancer incidence in five continents" volumes for several decades [6]. Cancer registry data in Russia were used for national reports and rarely for other research purposes, e.g., Saint-Petersburg PBCR published survival estimates. Participation in international projects restarted in 2015: CONCORD-2 international cancer survival study [7] included data from Arkhangelsk PBCR, CONCORD-3 studу included data from Arkhangelsk, Tomsk, Omsk and Karelia PBCRs [8], IARC "Cancer incidence in five continents" volume XI (2008-2012) included data from Arkhangelsk, Samara, Chelyabinsk and Karelia PBCRs [9]. IARC and World Health Organization (WHO-Europe) organized the first formal training course in cancer registration in Russia in 2015. NN Petrov National Medical Research Cancer Center completed the translation of ICD-O-3 into Russian in 2017 [10], and in 2018 the Ministry of Health recommended its use for PBCRs.

The current state of cancer registration
The cancer registration system in Russia can be formally described as passive. PBCRs collect information on all malignant and in situ neoplasms (according to the International Classification of Diseases, 10th revision, codes C00-96, and D00-09, respectively). More than 600,000 cases were registered annually starting from 2017 (640,391 including 83,752 non-melanoma skin cancer cases in 2019). Only new cases of invasive cancer (C00-96) are included in the national annual report. The basis of cancer registration is notification forms that physicians use to report cancer cases to the registries (Fig. 2). However, PBCRs can also use any available medical records to add information to the database. Cancer registration is organized according to the administrative division of the Russian Federation. All medical facilities, without exception, are required to notify the PBCR about newly diagnosed cases, and each regional PBCR is responsible for registering all cancers diagnosed in citizens residing in the region. It is unclear whether all regional hospitals and clinics, especially non-cancer and private, follow these instructions. Pathology labs do not report cases directly to PBCRs but send all information to clinicians responsible for notifying PBCRs. In cases of migration or death in different areas, PBCRs do not regularly perform information transfer.
PBCRs perform linkage using personal identifiers (name, date of birth, passport number, place of residence). Physicians are required to fill in notification forms summarizing clinical records for every cancer case. Multiple notifications can be provided for a cancer case with several hospitalizations or visits. Information from pathology laboratories is included in the forms and is not reported directly to the registry except for autopsy cases. Therefore, it is impossible to separate these two sources of information explicitly. PBCRs perform linkage with mortality records regularly (within the same region), but practices vary across the country.
According to Russian regulation from 1996, PBCRs send depersonalized cancer registry data from the regions to the Research Institutes of Oncology (now National Medical Research Cancer Centers). For example, registry data from the Northwestern Federal District (NWFD) regions are collected and maintained in NN Petrov National Medical Research Cancer Center, which, according to the law, is obliged to perform epidemiological analyses for various purposes.
Hematological malignancies, brain, and CNS tumors are often not referred to cancer hospitals in some regions, meaning that the registration of these cancer types could be incomplete [11]. In some regions, surgical treatment may also take place in general hospitals.
PBCRs report preliminary data annually in February/March and perform quality checks/traceback during the next six months. Thus, the previous year's data are considered complete in October/November of the following year. The annual national report includes information on incidence (absolute number of cases, crude rates, and age-standardized rates (ASRs) per 100,000 using Segi-Doll 1960 World population standard) based on aggregated data from the registry complemented by mortality data from the Russian Federal State Statistics Service (RSSS).
We obtained additional qualitative information about cancer registration procedures through personal communication with regional PBCRs of the NWFD. According to information acquired from the regions of the NWFD, PBCRs established the mortality linkage in all of them. However, it may be manual -cancer registry personnel or regional oncologists manually review death certificate records at regional civil registries to detect: 1) cancer patients from the registry who have died and 2) death certificates that mention cancer. PBCRs perform a semiautomatic procedure through manual searches in electronic databases of the state insurance system (Obligatory Medical Insurance) and civil registries. Most of the registries in the NWFD are based in the regional cancer hospitals (Oncology Dispensaries), except Saint-Petersburg and Kaliningrad, where registries are part of the regional medical statistics service. PBCRs have dedicated staff, but there is high regional heterogeneity in the number of employees in each registry. Most common software options for PBCRs have been developed in Moscow ("Cancerregister FB6") and Saint Petersburg ("NovelSPB Population-based Cancer Registry"). Some regions use their in-house software. All software tools follow national instructions but vary in the number of additional variables being processed. Trace-back procedures are not clearly described for cases initially registered with the information from death certificates (DCI), it is challenging to compare proportions of death certificate only (DCO) proportions as indicators of completeness of registration. DCO proportions are available from the national report for all regions (all cancers combined, including non-melanoma skin cancers), and with no sex decomposition. Overall national DCO proportion in 2019 was 5.3 % (range 0-14 % in different regions). National annual reports provide additional information of DCO% and morphologically verified cases proportion (MV%) for each cancer type (Table 1).

Comparison with international definitions and recommendations
We compared national instructions for cancer registration, listed in the corresponding national legislative document, to recommendations proposed by the European Network of Cancer Registries (ENCR) [12]. Cancer registration in Russia is consistent with some but not all ENCR recommendations (Table 2)  definitions and recommendations are outdated or missing compared to the most recent ENCR recommendations.

Cancer statistics reports
National cancer incidence and mortality reports are published annually and also available online since 2007. Cancer incidence from the reports is included in official state statistics. Forms with aggregated data collected every year from the regional PBCRs are the basis for the national report. There are sporadic local and international publication that use regional PBCR data [14][15][16].
PBCR data remain the primary source for cancer incidence, while cancer mortality statistics are based on the civil registry (Fig. 3). International reports and databases that include cancer statistics for Russia commonly acquire data from national sources (European Cancer Information System [17]), but some project or extrapolate data (e.g., GBD IHME). Comparison between all reports is warranted as both quality of primary data and model assumption may bias the estimates of cancer burden. Given the lack of full-scale quality assessment, we recommend caution in the interpretation of available national cancer statistics for Russia, especially for less common cancer types.

Conclusion
Russia has preserved a systematic approach to cancer registration at the national level. To our knowledge, cancer remains one of two disease entities registered in a population-based manner in Russia with individual-level databases (the second one is diabetes [18]). Cancer is also among only a few causes of death reported consistently across the country [19]. In 1999 new legislation introduced morphological coding, incidence date rules, and summarized processes to collect multiple variables in a population-based manner along with implementing electronic databases. However, it didn't provide an accurate description of quality control mechanisms. Lack of quality control means that apart from systemic discrepancies, we could also expect some regional heterogeneity in the data collection and processing. The other significant limitation of the current cancer registration system is that it is based on Table 1 Number of cases, deaths, incidence and mortality age-standardized rates per 100,000, proportion of cases registered with information from death certificate only (DCO %), proportion of morphologically verified cases (MV%) and mortality to incidence ratios for common cancer type for which information in both incidence and mortality is available (as reported by national annual report in 2019).  only two sourcesdeath certificates and clinical notification. This situation is typical for some Eastern European countries [20]. Pathology laboratories are a relatively centralized and essential independent source of information that PBCRs should not ignore, but in Russia, pathology labs do not directly send reports to PBCRs. In general, national legislation needs some updates. Full-scale implementation of ENCR/IARC recommendations, quality control procedures, and ICD-O-3 coding, along with proper training, should be prioritized for improving cancer registration in Russia. National training program for cancer registrars is not available in Russia, but it is highly recommended to improve data quality. Quality improvement requires dedication, human resources, and support from the local governing organs. However, investments in cancer registration are highly costeffective if we account for all possible applications of cancer registry data [21]. Further analyses are needed to assess the comparability (incidence dates, multiple primaries, etc.), validity (re-abstracting and recoding, the percentage morphologically verified (MV%) and cases reported by death certificate only (DCO%), missing information, internal consistency, etc.), completeness (historic data methods, independent case ascertainment, capture-recapture methods, etc.), and timeliness (comparison of database completeness at different time points) of Russia's PBCRs individual-level data [22,23].