RESEARCH

Health-systems efficiency in the Russian Federation: tuberculosis control

Efficacité des systèmes de santé en Fédération de Russie : la lutte antituberculeuse

Eficiencia de los sistemas de salud en la Federación de Rusia: control de la tuberculosis

Katherine Floyd^I; Raymond Hutubessy^I; Yevgeniy Samyshkin^II; Alexei Korobitsyn^III; Ivan Fedorin^IV; Gregory Volchenkov^V; Boris Kazeonny^VI; Richard Coker^VII; Francis Drobniewski^VIII; Wieslaw Jakubowiak^III; Margarita Shilova^IX; Rifat A Atun^{II, 1}

^IStop TB Department, World Health Organization, Geneva, Switzerland
^IICentre for Health Management, Tanaka Business School, Imperial College London, South Kensington Campus, London SW7 2AZ, England
^IIIWorld Health Organization, Moscow, Russian Federation
^IVMinistry of Health, Samara, Russian Federation
^VOblast TB dispensary, Vladimir, Russian Federation
^VIOblast TB dispensary, Ivanovo, Russian Federation
^VIIDepartment of Public Health and Policy, London School of Hygiene and Tropical Medicine, London, England
^VIIIDepartment of Infectious Diseases, Guy's, King's and St Thomas' Medical School, London, England
^IXResearch Institute of Phthisiopulmonology, Sechenov Moscow Medical Academy, Moscow, Russian Federation

ABSTRACT

OBJECTIVE: To conduct a comprehensive assessment of the case-mix of patients admitted to tuberculosis hospitals and the reasons for their admission in four Russian regions: Ivanovo, Orel, Samara and Vladimir. We also sought to quantify the extent to which efficiency could be improved by reducing hospitalization rates and re-profiling hospital beds available in the tuberculosis-control system.
METHODS: We used a standard questionnaire to determine how beds were being used and who was using the beds in tuberculosis facilities in four Russian regions. Data were collected to determine how 4306 tuberculosis beds were utilized as well as on the socioeconomic and demographic indicators, clinical parameters and reasons for hospitalization for 3352 patients.
FINDINGS: Of the 3352 patients surveyed about 70% were male; the average age was 40; and rates of unemployment, disability and alcohol misuse were high. About one-third of beds were occupied by smear-positive or culture-positive tuberculosis patients; 20% were occupied by tuberculosis patients who were smear-negative and/or culture-negative; 20% were occupied by patients who no longer had tuberculosis; and 20% were unoccupied. If clinical and public health admission criteria were applied then < 50% of admissions would be justified and < 50% of the current number of beds would be required. Up to 85% of admissions and beds were deemed to be necessary when social problems and poor access to outpatient care were considered along with clinical and public health admission criteria.
CONCLUSION: Much of the Russian Federation's large tuberculosis hospital infrastructure is unnecessary when clinical and public health criteria are used, but the large hospital infrastructure within the tuberculosis-control system has an important social support function. Improving the efficiency of the system will require the reform of health-system norms and regulations as they relate to resource allocation and clinical care and implementation of lower-cost approaches to case management for patients with social problems. Additionally, closer attention will need to be paid to the management of staff numbers in the tuberculosis system.

Keywords: Tuberculosis, Pulmonary/therapy; Hospitals, Special/utilization; Hospital costs; Hospital bed capacity; Bed occupancy; Hospitalization; Inpatients; Efficiency, Organizational; Health care surveys; Cross-sectional studies; Russian Federation (source: MeSH, NLM).

RÉSUMÉ

OBJECTIF: Conduire une évaluation globale d'un éventail de cas hospitalisés dans des établissements spécialisés dans la tuberculose et des raisons de leur admission dans quatre régions russes : Ivanovo, Orel, Samara et Vladimir. Nous avons également cherché à déterminer dans quelle mesure l'efficacité pourrait être améliorée en réduisant les taux d'hospitalisation et en restructurant les lits d'hôpitaux disponibles dans le système de lutte antituberculeuse.
MÉTHODES: Nous avons utilisé un questionnaire type pour déterminer de quelle façon les lits étaient utilisés et qui les utilisait dans les établissements spécialisés dans la tuberculose de quatre régions russes. Les données ont été recueillies afin de déterminer comment 4306 lits affectés à la tuberculose étaient utilisés et également concernant les indicateurs socio-économiques et démographiques, les paramètres cliniques et les raisons de l'hospitalisation de 3352 malades.
RÉSULTATS: Sur les 3352 patients étudiés, environ 70 % étaient des hommes ; l'âge moyen était de 40 ans et les taux de chômage, d'incapacité et d'alcoolisme étaient élevés. Près du tiers des lits étaient occupés par des patients tuberculeux à frottis positif ou à culture positive ; 20 % étaient occupés par des patients tuberculeux à frottis négatif et/ou culture négative ; 20 % étaient occupés par des patients qui n'avaient plus la tuberculose ; et 20 % étaient inoccupés. Si des critères d'admission cliniques et de santé publique étaient appliqués, moins de 50 % des admissions seraient justifiées et moins de 50 % du nombre actuel de lits s'avéreraient nécessaires. Jusqu'à 85 % des hospitalisations et des lits ont été jugés nécessaires lorsque les problèmes sociaux et d'accès aux soins ambulatoires étaient pris en compte au même titre que les critères cliniques et de santé publique.
CONCLUSION: Lorsque l'on utilise des critères cliniques et de santé publique, une grande partie de l'importante infrastructure hospitalières consacrée à la tuberculose en Fédération de Russie est superflue, mais cette infrastructure du système de lutte antituberculeuse joue un rôle de soutien social non négligeable. Il faudra, pour améliorer l'efficacité de la lutte antituberculeuse, entreprendre une réforme des normes et des réglementations du système de santé qui ont trait à l'allocation des ressources et aux soins cliniques et mettre en oeuvre des méthodes moins coûteuses de prise en charge des cas pour les patients présentant des problèmes sociaux. En outre, une attention accrue devra être portée à la gestion des effectifs dans le système de lutte antituberculeuse.

Mots clés: Tuberculose pulmonaire/thérapeutique; Hôpital spécialisé/utilisation; Coûts hôpital; Capacité lits hôpital; Occupation lit hôpital; Hospitalisation; Malade hospitalisé; Efficacité fonctionnement; Enquête système de santé; Etude section efficace; Fédération de Russie (source: MeSH, INSERM).

RESUMEN

OBJETIVO: Realizar una evaluación integral de los casos de pacientes ingresados en hospitales de tuberculosis y de las razones de su ingreso en cuatro regiones de Rusia: Ivanovo, Orel, Samara y Vladimir. Decidimos además determinar en qué medida podía mejorarse la eficiencia reduciendo las tasas de hospitalización y reclasificando las camas de hospital disponibles en el sistema de control de la tuberculosis.
MÉTODOS: Utilizamos un cuestionario estándar para determinar cómo estaban utilizándose las camas y quiénes las ocupaban en los centros de tuberculosis de las cuatro regiones rusas mencionadas. Se reunieron datos para determinar cómo se utilizaban 4306 camas de tuberculosis, así como sobre los indicadores socioeconómicos y demográficos, los parámetros clínicos y los motivos de hospitalización de 3352 pacientes.
RESULTADOS: De los 3352 enfermos encuestados, alrededor del 70% eran hombres; su edad media era de 40 años; y las tasas de desempleo, discapacidad y consumo abusivo de alcohol eran altas. En torno a una tercera parte de las camas estaban ocupadas por enfermos de tuberculosis con baciloscopía positiva o cultivo positivo; el 20% lo estaban por tuberculosos con baciloscopía negativa y/o cultivo negativo; un 20% estaban ocupadas por pacientes que ya no sufrían tuberculosis; y el 20% estaban libres. Aplicando criterios de admisión clínicos y de salud pública, menos del 50% de los ingresos estarían justificados, y se necesitaría menos del 50% del número actual de camas. Cuando se tuvieron en cuenta los problemas sociales y el escaso acceso a la atención ambulatoria además de los criterios antes mencionados, se estimó que hasta un 85% de los ingresos y las camas eran realmente necesarios.
CONCLUSIÓN: Gran parte de la amplia infraestructura hospitalaria dedicada a la tuberculosis en la Federación de Rusia es innecesaria si nos atenemos a unos criterios clínicos y de salud pública, pero la infraestructura de hospitales en el contexto del sistema de control de la tuberculosis cumple una importante función de apoyo social. Para mejorar la eficiencia del sistema habrá que reformar las normas y disposiciones relativas a la asignación de recursos y la atención clínica, y aplicar enfoques de bajo costo a la gestión de los casos de pacientes con problemas sociales. Además, deberá potenciarse la gestión de los trabajadores empleados en el sistema de atención de la tuberculosis.

Palabras clave: Tuberculosis pulmonar/terapia/utilización; Costos de hospital; Hospitales especializados; Capacidad de camas en hospitales; Ocupación de camas; Hospitalización; Pacientes internos; Eficiencia organizacional; Encuestas de atención de la salud; Estudios transversales; Federación de Rusia (fuente: DeCS, BIREME).

Introduction

The Russian Federation's health-care system is characterized by centralized planning and administration, government financing, and the provision of services to the public.¹ There are several vertical programmes and related infrastructure for diseases deemed priorities, including tuberculosis (TB), sexually transmitted infections, diabetes, vaccine-preventable diseases and HIV/AIDS. In 2001, there were 9.1 hospital beds per 1000 population, compared with 2.4 per 1000 for the United Kingdom and 4.1 per 1000 in the European Union.² An extensive primary care network comprising polyclinics, ambulatory care centres and feldsher stations also exists, but focuses on referral rather than gatekeeping and the management of patients. (Feldshers are medical/surgical practitioners who do not have full professional qualifications or status.)

Globally, the Russian Federation ranks twelfth in terms of the total number of new TB cases that occur each year.³ After a period of decline starting in the 1960s, the case-notification rate almost tripled during the 1990s, reaching 90.7 per 100 000 population in 2000.⁴ Similarly, the death rate from TB rose from 8 per 100 000 in 1991 to 20 per 100 000 in 2001.⁴ A large verticalized network of specialized institutes, dispensaries, hospitals, outpatient clinics, sanatoria and rural feldsher points exists to detect and treat cases; this network includes about 80 000 beds designated specifically for treating patients with TB. Box 1 provides further details of how TB control is implemented.^{5, 6}

With support from international and bilateral agencies, new laws are being adopted to gradually implement standard international practices in TB control.^7–9 Data from 2003 show that 25% of the population is covered by the internationally recommended control strategy known as DOTS.³ In 2003 a new regulation (prikaz) was implemented to enforce the use of standardized drug regimens; and in 2004 a prikaz was approved that introduced a recording and reporting system similar to that recommended in the DOTS strategy.^{10, 11}

Despite these developments, there is considerable scope for improving efficiency in the extensive TB-control system. Studies have shown that implementation of the DOTS strategy could substantially lower treatment costs for new smear-positive patients,¹² that inpatient care accounts for more than 50% of total TB-control costs,¹³ that health-system financing and provider-payment systems create perverse incentives for the use of lengthy and repeated hospitalizations (beyond those stipulated in the regulations),¹⁴ and that to compensate for inadequate social-support systems for TB patients, providers use sophisticated practices to enable lengthy admissions during winter.¹⁵ This contrasts with other countries that have a high burden of TB but where treatment is generally provided on an outpatient basis and it is unusual for patients to be hospitalized for long periods of time and large networks of TB hospitals do not exist.³

While these previous studies have identified frequent and lengthy hospitalizations and the factors that lead to these practices, they have not included a comprehensive assessment of the case-mix of patients admitted to TB hospitals and the reasons why patients are admitted: they have focused on new smear-positive or culture-positive patients in Group IA and not cases in Groups IB or II–VII (Box 1). Nor have they attempted to quantify the extent to which the existing number of beds and admissions could be reduced. To address these gaps, we conducted a detailed study of the utilization patterns of TB hospitals in four Russian regions. The study was designed to answer three questions. First: what kinds of patients are using the inpatient infrastructure of TB hospitals? Second: why are these patients being admitted? Third: what scope exists for reducing the use of the inpatient infrastructure in TB hospitals?

Methods

A cross-sectional bed census was conducted in TB facilities in four Russian regions: Ivanovo, Orel, Samara and Vladimir. These regions were chosen because collaborative links had been established among study investigators through projects funded by WHO, the United States Agency for International Development and the United Kingdom Department for International Development. The regions vary in terms of location, geographical size, population size, mean income level and number of TB-control facilities. Summary indicators for each region and the country as a whole are shown in Table 1.

A questionnaire was designed by a multidisciplinary study team; it was piloted in each of the four regions and refined through discussions with Russian and international experts (e.g., senior staff in specialist TB research institutes at federal level, chief TB doctors in each region where the survey was to be undertaken, WHO staff), including some that were bilingual. TB physicians in the four regions were trained to use the questionnaire. These physicians surveyed how all of the specialized TB beds were utilized on one day in Orel and Samara, and due to time and logistical constraints, collected data on large representative samples in Ivanovo (86% of all beds) and Vladimir (82% of all beds). The dates of the surveys varied for logistical reasons and to allow data to be collected for both summer and winter. Ivanovo was surveyed in September 2003 and the survey was repeated in March 2004; Orel was surveyed in September 2003; Samara was surveyed in November 2003; and Vladimir was surveyed in February 2004. The TB specialist in charge of each hospital ward was responsible for data collection. The patients' variables for which data were collected included socioeconomic and demographic indicators, clinical parameters, reasons for hospitalization and the classification of the case according to the Russian system in use in mid-2003 (Box 1).⁷

The cross-sectional study design required to understand the use of TB hospital beds is not compatible with collecting data on treatment outcomes; moreover such data were not necessary to answer our three study questions. Nonetheless, given the importance of treatment outcomes and the fact that they may vary according to whether treatment is provided on an inpatient or outpatient basis, this issue is addressed in the discussion section of the paper.

All data were entered into an Excel spreadsheet (version 7.0) and analysed in Excel and by using Stata software (version 8.0). Analysis was done in two stages. In the first, three major analyses were done: the socioeconomic and demographic data across regions were summarized and compared; bed utilization was quantified and compared across regions by identifying the percentage of beds occupied by each of the groups of patients in the Russian classification system and the percentage of beds that were unoccupied; and the frequency with which different reasons for hospitalization were reported were summarized and compared across regions.

In the second stage, we assessed the percentage of patients that would require admission, and the percentage of available beds that would be needed, for five sets of admission criteria. These admission criteria were defined in consultation with Russian experts and reflected clinical, public health, social and health-system justifications for admission. To assess the percentage of existing beds that would be required under each scenario, we assumed that the optimal average bed occupancy rate was 80%.¹⁶

Findings

A total of 3352 patients were surveyed, and the occupancy levels and reasons for utilization of 4306 beds were determined. Socioeconomic and demographic indicators for patients are summarized in Table 2. About 70% of patients were male; 13–19% were children; and 5–10% were elderly. The average age was 40. There were high rates of alcohol misuse among patients (32–49%), disability (17–39%) and unemployment (27–41%). In Ivanovo, 7% of patients were homeless. Rates of unemployment, disability, alcohol misuse, homelessness and the distribution of patients between rural and urban areas varied among the four study regions as did rates of multidrug-resistant TB.

The utilization of beds according to the Russian classification of cases is illustrated in Fig. 1. Around 25–30% of beds were occupied by patients in Groups IA or IB and 5–7% of patients were smear-positive or culture-positive. Overall, about 20% of beds were occupied by patients in Groups IA or IB who were not smear-positive or culture-positive (i.e., bacteriologically negative, BK-) at the time of the study, but numbers varied across oblasts (range = 7–23%). Patients in Groups II–VII (patients who no longer had active TB or who had extrapulmonary TB) also occupied about 20% of beds, but again numbers varied among oblasts (range = 18–32%). The numbers of patients with extrapulmonary TB within this group were small (data not shown). Overall, about 20% of beds were unoccupied (range = 8–37%). In every region at least one-third of beds were either unoccupied or occupied by patients who did not have active TB; in two regions (Orel and Ivanovo) the figure was > 50%.

When physicians were asked what was the single most important reason for hospitalizing a patient, the answers most frequently mentioned were that: the patient needed chemotherapy for TB, hospitalization was required according to existing regulations, the patient needed to be hospitalized to confirm or exclude the diagnosis of TB, the patient needed to be isolated to prevent transmission to others or to allow immunity to TB to develop, or the patient needed surgery (Table 3). Other reasons included on the questionnaire were rarely given as the main reason for admission (detailed data available from the authors).

The percentage of patients for whom admission would be required and the percentage of existing beds that would be needed according to different sets of clinical, public health, social and health-systems criteria are shown in Table 4. If clinical and public-health criteria were used, around 40–50% of patients would need to be admitted, and 40–50% of existing beds would be required (Table 4, scenario 1). If homeless patients were admitted to hospital in addition to those meeting clinical and public-health criteria (scenario 2) then about 50% of patients would require admission and 40–50% of beds would be needed. When alcohol misuse is added as a justification for admission (scenario 3), approximately 60% of patients would need to be admitted and 45–70% of beds would be required. When poor access to health services is added as a reason for admission (scenario 4), approximately 65–80% of patients would need to be admitted and about 50–80% of available beds would be needed. The averages across all regions are 77% of patients would need admission and 71% of existing beds would be needed. When clinical and public-health reasons, homelessness, lack of access to health facilities and an extended set of social reasons (patients with poor living conditions and/or no care at home) are all considered to justify admission, around 75–80% of patients would need to be admitted and 65–85% of existing beds would be needed. The average across all regions is 78% of patients would need admission and 76% of beds would be needed.

Discussion

Our study suggests that much of the Russian Federation's TB hospital infrastructure (80 000 beds or about 1000 per region) is not used according to clinical and public-health criteria, and a large share of hospital beds are used for social reasons and because patients have difficulty in accessing outpatient care. In the four regions that we surveyed, about half of admissions were justified and half of the beds would be required if clinical and public-health criteria were applied but up to 81% of admissions would be justified and 85% of beds would be needed when clinical, public health, social and health-system factors are considered.

Despite some differences in the demographic, geographical and socioeconomic characteristics of the regions and patients studied, the broad similarity across regions was striking in terms of the percentage of patients who would need admission when different criteria were used. This is in line with previous studies of TB control in the Russian Federation that have identified consistency in case management, costs and average length of hospital admission across regions as well as the important social-support function played by TB hospitals.^{9, 13–15} This consistency suggests that our results are likely to be applicable to the rest of the country's 88 regions.

If our results can be generalized, there is substantial potential for reallocating beds set aside for specialized TB care (that is, these beds could be used for other purposes or when this is not possible, fewer beds could be allocated for TB care), thus generating savings. In 1999, the cost of a bed-day for TB patients was consistently estimated at around US$ 3.00.13 If beds were used only for patients requiring admission for clinical and public-health reasons, more than half of the existing 80 000 beds set aside for specialized TB care (on average about 500 per oblast) could be reallocated, resulting in savings of around 15 million bed-days and US$ 45 million per year. For comparison, about US$ 245 million was spent on TB control in the Russian Federation in 2003,3 while in 2004 the government secured additional external funding from the World Bank and the Global Fund to fight AIDS, TB and Malaria (a combined worth of more than US$ 100 million over five years). Moreover, existing bed-day costs are part of a system in which salary levels are low (doctors' salaries are below the national average); there is inadequate funding for operating costs; and there is limited recent investment in buildings and equipment. Cost savings would be much higher if bed-day costs were more in line with the country's middle-income status and with the US$ 16.00 cost per day reported before the 1998 financial crisis.¹²

Of course, reallocation only makes sense if alternative services that both cost less and are at least as effective as the existing use of inpatient care can be developed for patients with social problems and poor access to health facilities and if cost-savings from the reallocation of specialist TB beds are available for reinvestment in such services. Evidence from other countries shows that outpatient care for TB patients is feasible, effective and costs less than inpatient care,^17–20 but in these settings social problems, such as alcohol misuse, are less common. Alternatives to inpatient care that have already been discussed in the Russian context include establishing shelter-type facilities that offer minimal clinical care, strengthening outpatient services and social support, and using isolation wards for treatment in general hospitals that have spare capacity.^{14, 21} Our findings suggest that assessing the costs and effectiveness of these alternatives and comparing them with existing care should be a priority. It is encouraging that data from recently introduced social-support programmes for TB patients in several Russian regions suggest that they can substantially improve treatment outcomes (from a treatment success rate of 65% to above the WHO target of 85%), with budgets of about US$ 20 000 per region per year (equivalent to the cost of operating 18 beds for one year).²²

As noted in the Methods section, our study design did not allow us to compare treatment outcomes for inpatients and outpatients, and it would have been difficult to do this since almost all patients are treated as inpatients. Nonetheless, our study demonstrates that even if some patients are being hospitalized because physicians believe this is the best way to ensure a good treatment outcome, many beds are unoccupied or occupied by patients who have already completed treatment. Two further limitations of our study are that the reasons treating physicians gave for hospitalizing patients were based on their own observation and judgement and that the data illustrate bed-use at only one point in time and seasonal variation may exist. We tried to overcome these limitations by investing time in training doctors to use the questionnaire and by collecting data during both summer and winter.

The success of infectious disease-control programmes is often determined by the constraints posed by the health systems in which they operate.^{23, 24} In the Russian Federation, there are six major barriers to the rationalization of the TB control system. First: hospital funding is based on the number of beds and bed occupancy. This preserves existing structures and provides perverse incentives for providers to maintain the number of beds and to hospitalize patients with TB.^{9, 14} Reallocating beds requires making changes to mechanisms of resource allocation and provider-payment systems for both TB control and the health system as a whole.¹⁴ Second: previously, regulations stipulated long periods of hospitalization for patients with TB.^{8, 9, 25–27} With the recent introduction of Ministry of Health Executive Order 109 these regulations have been formally abandoned but since the order does not explicitly specify new criteria for hospital admission previous regulations may still be used as a reference.¹⁰ Third: large numbers of staff are employed in TB-control facilities (17 860 doctors, 37 253 nurses, 18 515 hospital attendants and 18 500 ancillary staff). About one-third of the cost savings that could be made from reallocating beds is related to staff costs.¹³ Reducing or redistributing staff is politically difficult. Any reductions would need to be linked to improvements in pensions; current pension provision encourages staff to remain in their posts beyond retirement age. Fourth: outpatient services for TB care and social support for patients who are unemployed and/or misuse alcohol are poorly developed.^{9, 14, 15} Fifth: cost savings made from reallocating or reducing the number of beds would not necessarily be reinvested in TB control.¹⁴ Financing regulations (as in many other industrialized countries) prevent a shift of funds from health to social sectors and prevent a pooling of health-sector and social-sector budgets. For TB control to benefit from the reallocation of hospital infrastructure, and for such reallocation to be supported by existing stakeholders, it is essential that cost savings are reinvested in the TB-control system. For instance, funds could be used to provide incentives to staff to improve performance and outcomes and to strengthen weak outpatient and social-support services. The flexibility to reinvest savings would be facilitated by the introduction of global budgets for providers of TB services, which would provide flexibility to reallocate existing resources without a loss of income. Sixth: the prevailing experiences and attitudes of stakeholders (both staff and patients) may lead to strong resistance to change; this may be based on the view that inpatient care is necessary to ensure good outcomes and because patients expect to be hospitalized.²⁸

In the short term, these constraints mean that significant improvements in the efficiency of TB-control programmes that are based on optimal use of beds are unrealistic. However, in the medium term it should be feasible to positively influence stakeholders' attitudes, revise existing approaches of resource allocation, introduce more efficient provider-payment systems, implement phased reductions in the number of hospital beds and staff, change existing regulations governing the management of patients with TB, and strengthen outpatient care and social-support services. Encouragingly, a two-day workshop in Vladimir in 2004 with stakeholders from the federal and local levels, which was held to discuss the findings presented in this paper, indicated that there was a broad consensus on the need to gradually reduce the reliance on inpatient care and the steps that would be needed to achieve this.²¹

Our findings have important policy implications for TB-control systems in the Russian Federation and other countries of the former Soviet Union, where similar systems exist and where TB hospitals shoulder not only the cost of extensive periods of clinical care but also a substantial burden of nonclinical social support. Improving the efficiency with which existing resources are used will require health-system norms and regulations to be reformed as they relate to planning, financing and clinical care; this will take time. National and international agencies need to focus on medium-term to long-term improvements in the health-care systems in general rather than on short-term change.

Acknowledgements

We thank Dr Elena Kiryanova and Olga Medvedeva (Orel Oblast), Dr Tatiana Petrova (Ivanovo Oblast), Dr Tatiana Somova (Vladimir Oblast), Dr Galina Gusarova and Dr Sergei Kuznetsov (Samara Oblast), Dr Irina Danilova (WHO/Moscow), Dr Lyudmila Rybka (WHO/Moscow) and Mr Vladimir Skorobogatov (Ministry of Economy, Moscow), who assisted with the collection of data and commented on the data and text.

Funding: This study was funded by an unrestricted grant from the United Kingdom Department for International Development.

Competing interests: none declared.

References

1. Shishkin SV. [Reforms of financing of healthcare in Russia ]: Institute of Economies in Transition proceedings No. 25R. Moscow: Institute of Economies in Transition; 2000. In Russian.        

2. Tragakes E, Lessof S. In: Tragakes E, editor. [Health care systems in transition ]: Russian Federation. Copenhagen: European Observatory on Health Systems and Policies; 2003. p. 117-57. In Russian.        

3. WHO report 2005. Global tuberculosis control–surveillance, planning, financing. Geneva: WHO;2005. WHO document WHO/HTM/TB/2005.349.        

4. Shilova M. [Tuberculosis in Russia ]. Moscow: Research Institute for Phthisiopulmonology at the Sechenov Moscow Medical Academy; 2000. In Russian.        

5. Ministry of Health, Russian Federation. [Recording/reporting form number 47: information on health facilities infrastructure ]. Moscow: Ministry of Health; 2002. In Russian.        

6. Ministry of Health, Russian Federation. [Methodological guidelines for classification of persons registered at the tuberculosis dispensary facilities ]. Moscow: Ministry of Health; 1995 (Annex No. 5, Prikaz No. 324). In Russian.        

7. Maher D, Mikulencak M. What is DOTS? A guide to understanding the WHO-recommended TB control strategy known as DOTS. Geneva: World Health Organization; 1999. WHO document WHO/CDS/CPC/TB/99.270.        

8. Mawer C, Ignatenko N, Wares D, Strelis A, Golubchikova V, Yanova G, et al. Comparison of the effectiveness of WHO short-course chemotherapy and standard Russian antituberculous regimens in Tomsk, western Siberia. Lancet 2001;358:445-49.        

9. Coker RJ, Dimitrova B, Drobniewski F, Samyshkin Y, Balabanova Y, Kuznetsov S, et al. Tuberculosis control in Samara Oblast, Russia: institutional and regulatory environment. Int J Tuberc Lung Dis 2003;7:920-32.        

10. Ministry of Health, Russian Federation. [On improvement of TB control activities in the Russian Federation ]. Moscow: Ministry of Health; 2003 (Prikaz No. 109). In Russian.        

11. Ministry of Health, Russian Federation. [On implementing of TB recording/reporting system ]. Moscow: Ministry of Health; 2004 (Prikaz No. 50). In Russian.        

12. Migliori GB, Khomenko AG, Punga VV, Ambrosetti M, Danilova I, Ribka LN, et al. Cost-effectiveness analysis of tuberculosis control policies in Ivanovo Oblast, Russian Federation. Bull World Health Organ 1998;76:475-83 (Ivanovo Tuberculosis Project Study Group).        

13. Cost-effective TB control in the Russian Federation: report of the second project workshop, St Petersburg, Russian Federation, 18–20 June 2002. Moscow: WHO TB control programme for the Russian Federation; 2002.        

14. Atun RA, Samyshkin YA , Drobniewski F, Gusarova G, Skuratova NM, Kuznetsov SI, et al. Health system barriers to sustainable tuberculosis control in the Russian Federation. Bull World Health Organ 2005;83;217-23.        

15. Atun RA, Samyshkin YA, Drobniewski F, Kuznetsov SI, Fedorin IM, Coker RJ. Seasonal variation and hospital utilization for tuberculosis in Russia: hospitals as social care institutions. Eur J Public Health 2005;15:350-4.        

16. Gold MR, Siegel JE, Russell LB, Weinstein MC. Cost effectiveness in health and medicine. New York: Oxford University Press; 1996.        

17. Floyd K, Wilkinson D, Gilks C. Comparison of cost effectiveness of directly observed treatment (DOT) and conventionally delivered treatment for tuberculosis: experience from rural South Africa. BMJ 1997;315:1407-11.        

18. Floyd K, Skeva J, Nyirenda T, Gausi F, Salaniponi F. Cost and cost-effectiveness of increased community and primary care facility involvement in tuberculosis care in Lilongwe District, Malawi. Int J Tuberc Lung Dis 2003;7 Suppl 1: S29-37.        

19. Nganda B, Wang'ombe J, Floyd K. Cost and cost-effectiveness of increased community and primary care facility involvement in tuberculosis care in Machakos District, Kenya. Int J Tuberc Lung Dis 2003;7 Suppl 1:S14-20.        

20. Okello D, Floyd K, Adatu F, Odeke R, Gargioni G. Cost and cost-effectiveness of community-based care in rural Uganda. Int J Tuberc Lung Dis 2003;7 Suppl 1:S72-9.        

21. Improving the cost-effectiveness of TB control in the Russian Federation: the role of inpatient care. Geneva: WHO; 2005 (Policy Brief No. 2).        

22. Jakubowiak W, Danilova I, Bogorodskaya E, Borisov S, Malakhov K. Reducing default rates in Russia requires support for providers and patients. In: International Union Against Tuberculosis and Lung Disease conference abstract book. Paris: IUAT LD; 2004. p. S24.        

23. World Health Organization. The world health report 2003: shaping the future. Geneva: WHO; 2003.        

24. Walt G. WHO's World health report 2003. BMJ 2004;328:6.        

25. Ministry of Health, Russian Federation. [Guidelines on the forming of dispensary follow up groups ]. Moscow: Ministry of Health; 1995 (Attachment to the Ministry of Health and Medical Industry of the Russian Federation No. 324, 22 Nov 1995). In Russian.        

26. Ministry of Health, Russian Federation. [On approval of protocols of the treatment of TB patients ]. Moscow: Ministry of Health; 1998 (Prikaz No.33). In Russian.        

27. Ministry of Health, Russian Federation. [Methodological recommendations on implementation of the state guarantee programme ]. Moscow: Ministry of Health, Fund for Mandatory Insurance, Ministry of Finance; 2001. In Russian.        

28. Atun RA, Baeza J, Drobniewski F, Levicheva V, Coker RJ. Implementing WHO DOTS strategy in the Russian Federation: stakeholder attitudes. Health Policy 2005;74:122-32.        

(Submitted: 9 October 2004 – Final revised version received: 17 March 2005 – Accepted: 25 July 2005)

1 Correspondence to Dr Atun (r.atun@imperial.ac.uk).