While the numbers accessing antiretroviral treatment have steadily increased over time in sub-Saharan Africa there is growing concern over failures to retain patients in care. Professor Anthony Harries, a key player in establishing Malawi's national treatment programme, highlighted ten interventions to improve both access and retention during a symposium at the Fifth International AIDS Society conference in Cape Town in July.
From 2003 to 2007 antiretroviral treatment scale-up in sub-Saharan Africa increased from two percent to 30 percent. Progress in the public sector in Malawi was no less significant: in December 2003 nine sites were treating an estimated 3000 people and by December 2008 this had risen to a total of 170 sites providing treatment to close to 220,000.
However, this apparent success, Prof. Harries noted, is tempered by a growing concern over retention. He cited a study by Rosen et al demonstrating a steady decrease in retention in ART programmes in sub-Saharan Africa; at six months after the start of ART 79% were still in treatment, at 12 months the number decreased to 75% and by 24 months to 62%.
Treatment outcomes from Malawi mirrored this decline in retention: at 12 months 76% of those who had begun ART were alive and on ART; at 24 months 67% and at 36 months this had decreased further to 59%.
Drawing on experience from Malawi and research from other countries in sub-Saharan African, Professor Harries reviewed ways in which access and retention could be improved.
1. The creation and use of a simple, standardised monitoring and evaluation system
Identify new treatment starters on a quarterly basis. Cumulative cases can then be reported on (at 12 months and 24 months), and cohort analysis can be carried out at 12, 24 and 36 months to review programme performance. Outcomes can be divided into four categories:
- Those who are alive and still on ART
- Those who have died
- Those who have defaulted
- Those who have transferred out
This does not have to be a complicated system, Dr. Harries stressed. Simple treatment cards can be used to report on outcomes. They can be protected by cellophane sleeves and placed in high arch box files. On the left-hand side case findings can be noted and on the right treatment outcomes. Validation can be undertaken by a supervisor and a reward system initiated for good performance, for example, if cards are consistently filled out correctly.
2. Establish a national deaths registry
Many countries do not maintain national death registries. Accurate and consistent accounting of treatment outcomes is needed and can be categorised as: Death; default (a commonly recognised criterion is no clinic attendance for three months); stopped treatment; and transferred out.
Dr. Harries cited an example of ART patients in North Malawi who had been registered as defaulters. Of the 253 so registered 50% (127) had died, 8% (21) had transferred out but remained on ART, 15% (37) had stopped ART due to high costs of transportation and 68 (27%) could not be traced.
Dr. Harries noted that more can be done to actively trace those who miss appointments and/or those who have “defaulted”. He suggested tapping into community networks, making use of cell phones, verifying contact addresses and ensuring transfer procedures are correct.
3. Reduce death rates through active use of prophylaxis and TB case-finding
Dr. Harries cited the example of Malawi where as of December 2008 out of 215,449 who had been enrolled on ART, 66% were alive and on ART, 23,044 (11%) had died of whom 63% had died within the first three months. A further 11% (23,655) were lost to follow-up.
These numbers reflect a general trend in sub-Saharan Africa where death rates in the first 12 months are between 8% and 26%; people present late and are at increased risk. Undiagnosed TB, for example, is often the cause of death.
Dr. Harries suggested several interventions that would reduce the risk of early death including implementation of cotrimoxazole preventive therapy; active screening for TB in high-risk patients; active screening for cryptococcal antigen and the possible use of fluconazole prophylaxis and starting ART earlier. In Malawi he noted that 24% of eligible patients failed to begin ART after diagnosis.
Key in this process, Dr. Harries firmly believed, was the idea of a “Pre-ART clinic” to assist in preventing people being lost within the system, and to prepare patients for treatment.
4. Uninterrupted ART supplies
He noted that this will depend on accurate drug forecasting, timely placement of procurement orders and an efficient in-country distribution system. Given the time, cost and travel distances involved, if patients arrive at a clinic to find that medicines are out of stock, they are likely to question whether further attempts to return are worthwhile.
5. ART regimens that are non-toxic, simple (once-a-day) and free at the point of care
Dr. Harries underscored the need to replace the common first-line regimen (stavudine (d4T) / lamivudine (3TC) /nevirapine (NVP)) with the less toxic alternative of tenofovir (TDF) /emtracitibine (FTC) / efavirenz(EFV) in order to reduce the high rate of toxicity associated with d4T. However, even with this substitution questions remain as to how to manage potential renal toxicity associated with tenofovir, and treatment in pregnant women and children. As for free provision of treatment, studies have proven that providing ART free significantly decreases loss to follow-up.
6. Decentralise ART clinics and reduce the frequency of visits for stable patients
He strongly supported the creation of ART clinics in rural hospitals and health centres, along with task-shifting, to make ART services more accessible and easier to attend for follow-up visits, and cited the success of the Ministry of Health and Médecins sans Frontières (MoH-MSF) model of care in Thyolo, Malawi.
7. Linking ART services with the community
The Thyolo model demonstrated that with community support 95% of people remained alive and on ART compared to 76% who did not have community support. Dr. Harries noted that this clearly demonstrates that better treatment outcomes are achieved when close links with the community and its networks are established. These include: people living with HIV associations; home-care treatment programmes, family care support, referral systems for patients with side-effects; community assisted adherence counselling; and community-assisted defaulter tracing.
8. Reduce indirect patient costs.
Dr. Harries suggested looking at transportation costs for example, since frequent clinic visits may result in unaffordable costs for patients. Reducing the number of clinic visits and decentralising care also reduce the transport cost burden.
9. Using ART services to deliver other useful interventions
He cited, for example how ART can be delivered with anti-malaria treatment; the provision of nutritional support for the patient and his/her family; incorporating family planning as well as combined clinics for ART, diabetes mellitus (DM) and hypertension (HT).
10. Thinking out of the box
Dr. Harries proposed the creation of an infrastructure that would support lifelong adherence. He suggested, for example, twice-daily radio announcements reminding patients to take their medications or having ART stations at the workplace as well as pre-packaging ART for collection at school or from the local shop. This would necessitate the appropriate research to support it.
Retention is just one more vital component of any continuum of care if treatment outcomes are to be improved and sustained together with the provision of ongoing resource and financial support.
Harries, A. D. Improving access and retention in ART programmes. Fifth IAS HIV Conference on Pathogenesis, Treatment and Prevention, Cape Town, South Africa, Symposium, July 2009
Zachariah R et al. Payment for antiretroviral drugs is associated with a higher rate of patients lost to follow-up than those offered free-of-charge therapy in Nairobi, Kenya. TRSTMH, 2008. (freely available here).
Mermin J et al. Cotrimoxazole preventive therapy (CPT) and insecticide treated bednets in the reduction of malaria episodes in Uganda. Lancet 367, 1256-610, 2006.
Massaquoi et al. Probability of attrition at district hospital and health centres, Thyolo, Malawi. TRSTMH 103 (6): 594-600, 2009. (freely available here