Hospital engagement lessons from the five-country WHO/CIDA initiative 2009-2013 Knut Lönnroth, Mukund Uplekar, Monica Dias, Diana Weil WHO/GTP/PSI On behalf of all project country teams
Project objectives 1. Test and scale up approaches for early detection of TB: a) Hospital engagement (public and private) b) Contact investigation c) Screening in high risk groups (PLHIV, people with diabetes, urban slums, etc) 2. Improve early detection of TB in countries with large estimated case detection gaps 3. Contribute to the global evidence base Main focus on large hospitals in urban areas Specific target: "to detect an additional 28 500 cases in the five countries". Funding from CIDA (Canada)
Countries Democratic Republic of Congo (Kinshasa, 20 hospitals) Ghana (Accra and Kumasi, 10 hospitals) Philippines (Manila, 17 hospitals) Swaziland (Whole country, 35 hospitals/clinics) Vietnam (Hanoi, Hue, Ho Chi Minh City, 3 hospitals)
Preparations 2009-10 Desk study on case detection gaps and missed opportunities for early detection Baseline assessment, including all selected project hospitals: TB screening, diagnosis, treatment, referral and notification routines Estimated case load in each project hospital Status of contact investigation and other risk groups screening Implementation plan, budget and target setting in each country M&E framework with core indicators for all countries
Hospitals baseline situation Competent staff, equipment, high case load Large number of patients tested and diagnosed with TB (exact numbers were hard to get) Non-standardised and heterogeneous screening and diagnostic approaches CXR widely used, no clear guidelines (for screening or diagnosis) Many with CXR abnormalities not undergoing bacteriological testing No/limited quality assurance of labs
Baseline, cont. Lack of standard treatment protocols, no registers, large default? (no data) No "DOTS corner" or TB coordinator or internal referral routines No formal notification routines: Incomplete data on contribution of hospital notification to case notification No clear linkages with peripheral centers for referral and follow-up Weak regulation/policy about hospitals public health responsibilities Financial disincentives for following NTP guidelines User charges - cost to patients depend on insurance status Thus: the old PPM problem statement!
Interventions 2010-2013 1. Dialogue between NTP, hospitals and MoH (including department responsible for hospitals) and WHO 2. Agreements on role division, diagnostic routines and notification and referral mechanisms 3. Reinforcement of policy to identify people with suspected TB: Instructions to OPD staff "Screening officers" in Ghana and Swaziland Agreement with radiology departments on the approach of identifying people with suspected TB among people with a chest radiography
Interventions, cont. 4. Upgrading of laboratories, placement of lab technicians as required 5. Establishment of a hospital DOTS coordinator / team and DOTS corner 6. Training of all relevant staff 7. Establishment of routine recording and reporting practices 8. Supervision, M&E routines
Example: The Philippines
Impact on testing and diagnosis 90000 80000 0.1 0.09 Number of bacteriological tests done Number 70000 60000 50000 40000 30000 20000 10000 0 2010 2011 2012 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 Number of bacteriologically positive cases diagnosed Proportion of OPD attendees with bacteriological test for TB (%) Proportion bacteriologically positive out of all tested (%)
Impact on case notification from Number of cases notified 14000 12000 10000 8000 6000 4000 2000 0 project hospitals Implementation starts Cumulative: >29 000 cases 33% 49% 56% 2007 2008 2009 2010 2011 2012 Number of bacteriologically positive cases diagnosed Total number of TB cases diagnosed and notified %: proportion bacteriologically positive cases of total cases notified
Impact on overall notification? Manila, Philippines Cases per 100,000
Initial default, referral, feedback after external referral, treatment outcomes Focus on referral for treatment, only ambulatory treatment by hospital for people living close to hospitals (NTP drugs). Treatment success rate of those started on treatment via DOTS corner same as other NTP patients (>85%), where disaggregated (Philippines and Ghana). Reduced "internal and external leakage" after documented large problems in the Philippines and Vietnam: Initially large "internal leakage" (diagnosed but not referred to DOTS corner) Low proportion with feedback after referral for treatment at district level NTP facility
Improving internal referral rate in Philippines (% of ss+ to DOTS corner of all diagnosed) 100% 90% 80% 87% 70% 60% 51% 57% 50% 40% 31% 30% 20% 10% 0% 3rd-4th Q2010 1st-2nd Q2011 3rd-4th Q2011 1st-2nd Q2012
Improving feedback on referral in Vietnam
Remaining referral challenges Large catchment areas difficult to refer to distant districts/provinces Mobile populations in urban areas what is the "home district"? Some want to stay with "specialist" / private sector NTP district facilities not fully sensitized about new referral policies incomplete feedback
Success 1: Better surveillance! Notifications from hospitals increased substantially Timely information about newly diagnosed cases and referrals Much better surveillance of initial default and treatment outcomes.
Success 2: Improved quality of diagnosis, referral and treatment! Standard procedures for identifying people with suspected TB EQA of labs Substantial increase in bacteriological testing and proportion bacteriologically confirmed cases Formal links with radiology department towards systematic bacteriological evaluation of people with abnormal CXR Improved internal and external referral Good treatment outcomes
Increased, earlier case detection? Notification from hospitals increased substantially Overall notification trends variable, background epi trend unknown: hard to interpret changes! Small proportion of all hospitals engaged so far! Maybe limited additional number of cases? But probably earlier diagnosis - better infection control
Critical components for success Political commitment to engage hospitals for public health service quality and impact (with financing and clear directives that are stronger than financial disincentives) Dialogue between NTP, MoH department dealing with hospitals, hospitals managers and clinicians (=time and patience) Dedicated and accountable staff for internal and external coordination The rest is basic "DOTS"
Next steps (beyond the project) Scale up to all hospitals (domestic funding, GF, etc) Philippines: By 2015, engage 90% of all the public hospitals and 65% of the private hospitals Vietnam: By 2014, 39 additional hospitals. Swaziland: Almost whole country already covered Ghana and DRC: Beyond the capital cities Use and interpret CXR correctly and introduce quality control Use of rapid molecular testing (link hospital engagement to Xpert scale up) Review definition of "suspected TB" to improve early detection Use standard M&E data to further improve referrals and treatment follow-up Further epidemiological assessment of impact on case detection and notification
Thanks!