Hans V. Hogerzeil Director

Transcription

1 ESSENTIAL Medicines MONITOR Issue 1* (2009) We are back new name, same concept = Hans V. Hogerzeil, Director, Department of Essential Medicines and Pharmaceutical Policies, Richard Laing and Kathleen Hurst, Editors e are very pleased to W re-launch the Essential Drugs Monitor. In keeping with the changing times, it has a new name and a new format, but it continues its 25-year tradition as a major advocacy and information tool for the essential medicines concept, and how it is being implemented in practice. In 2006, we decided that it was time for another survey of Monitor readers to find out how you wanted to see the journal develop (see p.3 for more details). Since that time, the WHO Medicines Department has been restructured and the way we relay information about our work has been reviewed. This resulted in the creation of a new unit, Medicine Information and Evidence for Policy. The unit brings together IN THIS ISSUE: Hans V. Hogerzeil reveals more about the Essential Medicines Monitor. Kiki de Jong gives an overview of results of our readership survey. John Chalker and colleagues provide an insight into one of the most important public health issues of our times adherence to antiretroviral therapy. Hans V. Hogerzeil Director the Essential Medicines Monitor, and the Medicines Documentation Centre with the electronic Medicines Bookshelf, available on the web at: en/ and on CD on request. What can you expect from the new Essential Medicines Monitor? an electronic A4 document, keeping the EDM look, with two three articles of approximately1300 words each, published 10 times per year. The aim is eventually to make the articles available in all six WHO official languages, for which we may have to rely on volunteer support articles will be posted on the Medicines web site together with Richard Laing Editor Kathleen Hurst Editor selected correspondence received about the articles for those with limited web access who may have problems downloading articles, please let us know and we shall try to send you a pdf of the articles four times a year a Published Lately feature will be posted to highlight WHO and other relevant publications and a Dates for your Diary section will give information on forthcoming courses and conferences approximately once every two years a compilation of the articles will be printed and distributed in hard copy to readers in developing countries who request it. The first issue will appear by the end of 2009.

2 2 ESSENTIAL MEDICINES MONITOR RATIONAL USe The Monitor will continue to bring an international perspective to medicines issues, focusing on developing and transitional countries, stimulating discussion and sharing news and experiences on core areas of work. We encourage the submission of articles from WHO field staff working on medicines issues, and from our partners in WHO Collaborating Centres, nongovernmental organizations and others working in the area of essential medicines. Our aim has been to devise a flexible and sustainable plan for a journal that is perfectly attuned to WHO s mission to promote access to quality-assured essential medicines within national health systems. The Monitor will once again allow the voice of the essential medicines family to be heard, to promote equity and social justice, in line with the renewed emphasis on primary health-care values. Finally, our thanks to the hundreds of readers who took the time to respond to the readership survey. We know that many people have been waiting for news of the Monitor, we hope that you approve of our plans and look forward to receiving your comments. Please send these to: emmonitor@who.int *The Essential Medicines Monitor has evolved from the Essential Drugs Monitor, the last issue of which - number 34 - was published in 2005, prior to our readership survey. You tell us what you think: results of the Monitor s readership survey = Kiki de Jong questionnaire was sent A out to readers with the last edition of the Monitor asking your opinions of the publication. I am a Masters student in International Public Health at the Vrije University, Amsterdam, and during an internship at WHO, I analysed the results of these questionnaires. I also approached some respondents to do an in-depth telephone interview with me, to obtain additional information. Here I give an overview of the results. records). Kiki de Jong What did we learn? We gained a much better picture of the work and needs of you, our readers. We learned that you consider the Monitor unbiased; it is useful in your professional life; and it contains information not available elsewhere. We discovered that the Monitor s subscribers only reflect a small percentage of the actual number of readers, as on average there are six readers for each copy. In addition, the Monitor is copied for use at workshops and seminars, which also increases the readership. The Monitor is often shared with students or colleagues, some working in the field of pharmaceuticals and others not. The majority of readers who responded live in Africa and South- East Asia. Fifty-nine per cent are either doctors or pharmacists, 4% are nurses and 4% teachers. The remaining 33% represent a wide professional spectrum ranging from public health administrators and scientific journalists to government officials. Readers work covers policymaking, management, education and training, communications and all aspects of primary health care. Many people (49%) read the Monitor for educational purposes and 30% because it helps them with general policy-related work. Some also used the Monitor to obtain ideas for research or used it as background information. The Monitor is often used to train staff, students and the public with earlier issues of the journal often kept for reference. Interestingly some readers translated the Monitor into their own local languages for wider dissemination. It seems that you think the level of language of the Monitor is acceptable as it is and that it should not be more technical. Some of you want to see more photographs and charts included, but others disagree, because the size of the journal would

3 ESSENTIAL MEDICINES MONITOR 3 increase. Most of you read all or almost all of the Monitor, which is very encouraging for us to hear. Many respondents asked for the Monitor to appear more frequently. Others state that they are not receiving issues regularly. A male doctor from the Western Pacific region said: It might be an endangered species, but still... It is my pleasure to receive the Monitor, when I almost forgot it existed. The greatest problems seemed to be in Africa and South-East Asia, and a doctor from the Americas said: Please keep printing, computer/ Internet access is a convenient myth in our part of the world. Some people could access the Internet but preferred the paper version. What now? very clearly how much you use the Monitor and some of you in remote areas even said that it was the only source of information you had about medicines. Write and tell us what is happening in your country or professional area to promote the essential medicines concept, how you are tackling problems and the lessons you have learnt along the way. Whenever we can, we will share this information with other readers. emmonitor@who.int As you see from our editorial, we share your wish for a more regular and frequent publication. Some readers would prefer an A4 format to facilitate photocopying and filing. However, other readers prefer the Monitor to stay A3, since they like the format and feel it gives the Monitor its own identity. The public education, national medicines policy and rational use sections interested you most. We received hundreds of suggestions of subjects you would like to see appear in future issues. Many of you want more specific country coverage. Some readers asked for more information about traditional medicines, rational use and safety of medicines. You also asked for more input from readers. In the past, there has been a section for letters to the editor, and we want to revive this and provide a lively forum for discussion on essential medicines issues worldwide. Your comments are very important to us. Because the cost of distributing hard copies of the Monitor is so high, we asked your opinion of the electronic version. Unfortunately, many readers do not have (good) access to the Internet and are not able to consult the electronic version. In this issue s editorial, we give more details of how we plan to make the Monitor once again a regular feature of your working lives. You told us Finally, thank you all, once again, for your interest in the Monitor, your encouragement and your ideas. Figure 1. Respondents by profession Nurse 4% Teacher 4% Figure 2. Respondents by type of work institution Academic/research institution Government department NGO Private sector Intergovernmental organization Other 33% Other Doctor 33% Pharmacist 26%

4 4 ESSENTIAL MEDICINES MONITOR Editor s note: Adherence to antiretrovirals is critical for longterm effective therapy with first-line medicines. This important study describes how groups from the International Network for the of Drugs in East Africa developed and tested simple facility-based indicators to measure adherence. The report below demonstrates that such measurements can be reliably performed, meaning that adherence defined in different ways can be routinely reported. This work also demonstrates the value of operational field research carried out by national researchers and health workers within a network. Developing standard methods to monitor adherence to antiretroviral medicines and treatment defaulting in resource-poor settings The International Network for the of Drugs Initiative on Adherence to Antiretroviral Therapy = John Chalker, Tenaw Andualem, Hailu Tadeg, Lillian Gitau, Joseph Ntaganira, Celestino Obua, Paul Waako T he 2004 International Conference on Improving Use of Medicines highlighted the urgent need to develop strategies to improve adherence to antiretroviral treatment (ART) ( Accepted wisdom is that if the ART adherence rate is less than 90 95%, treatment can fail, and the virus may become resistant. 1,2 A review of adherence studies for chronic illnesses found that achieving adherence rates above 80% is difficult, even in resource-rich countries. 3 Therefore, the ability to accurately monitor adherence rates for ART and immediately address problems is crucial. Although many countries are scaling-up ART programmes, no one has developed any practical approaches to monitor treatment adherence programatically. The International Network for the of Drugs Initiative on Adherence to Antiretroviral Therapy (INRUD IAA) is taking on the challenge. INRUD, comprising 25 member groups, was established in 1989 to design, test and disseminate effective strategies to improve the way medicines are prescribed, dispensed and used, especially in resource-poor countries. In 2006, national AIDS control programmes and INRUD groups carried out surveys in Ethiopia, Kenya, Rwanda, Uganda and the United Republic of Tanzania to assess how ART programmes were tracking patient adherence and treatment defaulting. Findings showed that programmes defined treatment adherence and patient defaulting differently the surveys identified 14 different definitions of defaulting, ranging from missing an appointment by one day to six months. As a result, reliable comparisons were impossible. Yet, although data collection and measurement were haphazard, clinics and pharmacies were recording much useful information. Developing indicators through collaboration Representatives from the INRUD g roups and AIDS cont rol programmes that had coordinated the survey and technical staff from Management Sciences for Health, WHO s Department of Technical Cooperation for Essential Drugs and Traditional Medicine, the Karolinska Institutet, and Harvard Medical School gathered in Entebbe in April 2006 to discuss the survey findings. The participants agreed on the need to standardize definitions of adherence and defaulting and define practical measurement methods. Standard indicators could help identify patients at risk, monitor facility and programme performance, and track changes. A measurement methodology would need to be practical, affordable and reproducible in any setting and provide reliable results. Participants drafted core indicators to measure treatment adherence and defaulting. They also suggested complementary indicators to address causes of good and bad adherence. The set of core indicators and their data sources are: S e l f - r e p o r t e d d o s e s o f antiretroviral (ARV) medicine missed over a recent period (interviews or clinical records). Number of days that ARV medicines were dispensed over the last six months or a year (pharmacy records). P a t i e n t a t t e n d a n c e a t appointments and the number of

5 ESSENTIAL MEDICINES MONITOR 5 A photo opportunity gives participants a break from discussing country survey results at a meeting in Entebbe. Photo INRUD days until reappearance following a missed appointment (clinic appointment logs). Pill counts at each patient s clinic visit compared to their expected pill consumption (clinical or pharmacy records). Self report-based adherence measures A clinician or pharmacist can easily collect data for this indicator by asking patients whether they have missed any doses of pills in a number of days, and if so, how many. Using clinical records to measure this indicator is possible only if the question has been asked consistently and recorded routinely. In practice, clinicians or pharmacists may have asked patients about their adherence but not recorded the answer. Also, the recall period may have ranged from their adherence yesterday to that since the last clinic visit. Three indicators could be measured using self-reporting: Percentage of patients with full adherence to ART (i.e., no doses missed in the recall period, which is 3 days in the INRUD IAA methodology). Percentage of ARV doses patients took during the recall period. Percentage of patients with more than 95% adherence to ARV treatment (i.e., missed no more than 1 dose in 20, which for periods shorter than 3 days, is equivalent to full adherence). Dispensing-based adherence measures Pharmacy dispensing records are useful to measure longer-term adherence patterns. By counting the number of days that medication is dispensed over a period, three other adherence indicators can be calculated: long-term adherence in ARV use, the achievement of longterm adherence targets, and the rate of discontinuing ARVs. Typically, this methodology is based on abstracting computer databases, such as those from health insurance companies. The INRUD IAA project tested the feasibility of sampling written records. The dispensing-based adherence measures are defined as: Average percentage of days a sample of patients received ARVs for a defined period, such as 6 months (180 days) (e.g., patients at a facility receive ARVs an average of 162/180 days or 90% of days). Percentage of sampled patients who received ARVs for a target percentage of days in a defined period (e.g., 90% of patients receive ARVs for at least the 95% target over 6 months 171/180 days).

6 6 ESSENTIAL MEDICINES MONITOR Percentage of patients who experienced a gap in ARV availability of more than 30 days in a row during a defined period. Using dispensing data may overestimate true adherence the patient may have received the medicine, but did not consume it correctly. However, if the patient never received the medicine, then they cannot adhere to treatment; for example, if the patient was dispensed medicine for 145 out of 184 treatment days, then the patient s maximum adherence rate could only be 79%. Pill count-based adherence measures Pill counts are used by some ART programmes to compare a patient s actual and expected consumption since the pharmacy last dispensed the medicine. If records include pill counts, the data can be used to calculate the pill count adherence measures. Patient attendance-based defaulting measures A missed appointment should trigger programme action to reach out to patients at risk of defaulting on their treatment; however, because the patient may have had extra days of medicine, attendance failure within three days of an appointment can also be a trigger point. Finally, the programme s failure to re-establish contact with patients within 30 days of a missed appointment indicates the level of treatment defaulting. The three performance indicators related to defaulting could be: Percentage of patients who do not appear for their appointments on the scheduled day. Percentage of patients who miss their original appointment, but come within 3 days of the missed appointment. Percentage of patients who miss their appointment and who do not reappear at the clinic within 30 days of the missed appointment. Complementary indicators Complementary indicators can help identify why patients have problems adhering to treatment; for example, staff with high average workloads may not have the time to adequately counsel patients. Data for these indicators can be collected at the same time as data for the core indicators. Complementary indicators and their data sources include: Average staff workloads (for example, patients per hour); percentage availability and stockouts of key ART-related medicines (interviews and record review). Percentage of prescribed medicines that were dispensed; percentage of patients who know how to take their medicine; percentage of medicines properly packaged and labelled; clinical or functional status, such as whether they are able to carry out normal activity (patient exit interviews). Age; gender; tuberculosis status; WHO disease stage at initiation of ARVs; CD4 count or viral load at treatment initiation and in the last 6 months (clinical or pharmacy records). Field testing methods Ethiopia, Kenya, Rwanda and Uganda tested the feasibility and reliability of collecting the adherence indicators between October 2006 and June The results of the surveys are published elsewhere. 4 The sampling strategy included 20 health facilities with at least 100 patients on ARVs chosen randomly in each country. Data collectors were practicing pharmacists, doctors, or senior-level students. Teams of three, four or five data collectors surveyed a John Chalker Tenaw Andualem Hailu Tadeg Lillian Gitau

7 ESSENTIAL MEDICINES MONITOR 7 single facility in one day and entered the day s data in the evening. In each facility, data collectors randomly sampled medical and pharmacy records and interviewed 30 patients as they exited the clinic. In the first two field tests (Kenya and Rwanda), data collectors selected two retrospective cohorts each comprising 100 patient records. The first cohort included patients who had attended the clinic during the month one year before (Kenya) or six months before (Rwanda). Data collectors examined the pharmacy records to see how many days of medicine had been dispensed over the period. The second cohort was used to track patients from the previous three months to see when and if they showed up for their next appointment. Pill count and self-reported adherence data were included if mentioned in the records. In Uganda and Ethiopia, data collectors combined these two samples into a single cohort of 100 patients who had attended the clinic 6 months previously. Results and discussion Self-reported adherence in exit interviews Interviewers carried out 1,631 interviews in the four countries, averaging 20 per facility. Some facilities in each country were surveyed on days with few patient appointments. Therefore, confirming the number of patients expected on the day of the survey is very important in planning. In nearly all facilities, most patients claimed full adherence over the last 3 days however, some did not. These individuals are most at risk of poor adherence, which is best captured by this indicator: Percentage of patients who claimed full adherence to ART over the last 3 days. Consistently using a standard question to assess individual patients recent adherence can rapidly identify patients who need counselling. On a larger scale, this information can help a programme manager focus on facilities having problems maintaining high levels of treatment adherence. Dispensing-based adherence measures In almost all facilities it was possible to assess the number of days of ART dispensed over the last 6 months. More than 6,500 records showed that it was possible to calculate: Percentage of days that a sample of patients received medicines over the last six months. Percentage of patients with gaps in treatment of more than 30 days over the last six months. Monitoring data on long-term adherence could be a useful tool to identify patients with adherence problems. Aggregate data would help a programme manager identify facilities with problems in drug supply, dispensing, or patient adherence in addition to highlighting facilities that are performing well. The data would also be useful for monitoring facility performance over time. Clinics with computerized dispensing systems would find this monitoring particularly easy. Defaulting Almost all facilities recorded the dates of attendance and next appointment, so that calculating the two attendance indicators was nearly always possible: Percentage of patients who arrived on or before the day of their next appointment. Percentage of patients who arrived within 3 days of their appointment. A standard approach for identifying patients who miss appointments can help programmes develop community outreach systems. In Joseph Ntaganira Celestino Obua Paul Waako

8 8 ESSENTIAL MEDICINES MONITOR addition, the information would allow a programme manager to help facilities with a patient attendance problem identify causes and appropriate interventions. Usefulness of pill counts and self-report in clinic notes Overall, only 15% of 6,551 patient records included a pill count; therefore, calculating adherence measures based on pill counts in medical and pharmacy records does not appear to be widely applicable. More records included a selfreport adherence measure (45% overall), although this measure was infrequently recorded in Rwanda (10%). However, the methods used to derive these self-report measures varied, which makes comparisons problematic. In Ethiopia, for example, the method of recording self-reported adherence was to use a G to indicate better than 95% adherence (good) or an F (fair) or a P (poor) for less than 95%. Of the 83% of records that included a self-report measure, 96% were rated good. Conclusion The INRUD IAA field tests examined four categories of indicators for adherence to antiretroviral medicines and treatment defaulting: 1) self-reported adherence from exit interviews, 2) days supplied by medicine, 3) patient attendance, and 4) pill counts and self-reports in clinic records. The first three methods offer feasible approaches to standardizing measures of adherence and defaulting in lowresource settings. Pill counts are used too infrequently; whereas, selfreports in clinic records appear more promising. However, the consistency of the data-gathering methods needs to be assessed. The four field tests provided excellent evidence that these indicators can be measured in resource-poor settings. Managers will be able to examine the causes of poor performance in certain facilities and work with them to make improvements for example, by promoting both community involvement and individual patient ownership and responsibility, which proved successful in a South African township. 5 Facilities that are doing well can also share lessons on how to achieve exceptional performance. However, only by monitoring adherence and defaulting can we know where and what kind of interventions are needed. The INRUD IAA has now validated these core adherence indicators against clinical outcomes so they can identify facilities needing attention. 6 It is developing a manual with supporting software that describes the recommended indicators, data collection methodology, and analyses, similar to WHO s How to investigate drug use in health facilities. 7 WHO will also publish this adherence indicator package when it has been field tested. Finally, the participating countries are using qualitative methods to investigate the reasons for good and poor adherence at both patient and facility level as a step towards developing appropriate strategies for improvement. References 1. Arnsten JH et al. Antiretroviral therapy adherence and viral suppression in HIV-infected. Clinical Infectious Diseases, 2001;33: Paterson DL et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV/AIDS. Annals of Internal Medicine, 2000;133: DiMatteo MR. Variations in patients adherence to medical recommendations: a quantitative review of 50 years of research. Med Care, 2004;42: Chalker J et al. Measuring adherence to antiretroviral treatment in resource-poor settings using available routine data: a health system approach. Submitted for publication. 5. Kasper T et al. Demystifying antiretroviral therapy in resourcepoor settings. Essential Drugs Monitor 2007;32: Ross-Degnan D et al. Indicators of adherence to antiretroviral therapy based on routine data in treatment programs in Africa: a validation study in four countries. (Submitted for publication). 7. How to investigate drug use in health facilities: selected drug use indicators. Geneva: World Health Organization; The authors would like to acknowledge the other members of the INRUD Initiative for Adherence to Antiretroviral Medicine (INRUD-IAA): Ethiopia: Abraham Gebre Giorgis, Ethiopian Drug Administration and Control Authority; Gabriel Daniel, Management Sciences for Health. Kenya: Michael Thuo, Management Sciences for Health, Dorine Kagai, the Kenya National AIDS and Sexually Transmitted Infection Control Programme. Rwanda: François Ndamage, Treatment and Research AIDS Centre/Ministry of Health. Uganda: Robert Balikuddembe, Makerere University Medical School. Harvard Medical School Drug Policy Research Group: Dennis Ross-Degnan and Anita Wagner. Division of International Health of the Karolinska Institutet: Göran Tomson, Rolf Wahlström, Stefan Peterson. WHO: Richard Laing. Management Sciences for Health: Keith Johnson, Lloyd Matowe, Jude Nwokike. With thanks to all patients and staff in the respective health facilities. Corresponding author: John Chalker, Center for Pharmaceutical Management, Management Sciences for Health, 4301 North Fairfax Drive, Suite 400, Arlington VA 22203, USA: ( INRUD@msh.org); Tenaw Andualem and Hailu Tadeg, Management Sciences for Health, Addis Ababa, Ethiopia; Lillian Gitau, Center for Drug Management and Policy, Sustainable Health Care Enterprise Foundation, Nairobi, Kenya; Joseph Ntaganira, Department of Epidemiology and Biostatistics, School of Public Health, National University of Rwanda, Butare, Rwanda; Celestino Obua and Paul Waako, Department Pharmacology and Therapeutics, Makerere University Medical School, Kampala, Uganda.