Evaluation of Two Strategies for Community-Based Safety Monitoring during Seasonal Malaria Chemoprevention Campaigns in Senegal, Compared with the National Spontaneous Reporting System

Evaluation of Two Strategies for Community-Based Safety Monitoring during Seasonal Malaria... Pharm Med (2018) 32:189–200 https://doi.org/10.1007/s40290-018-0232-z OR IGINAL RESEARCH ARTIC L E Evaluation of Two Strategies for Community-Based Safety Monitoring during Seasonal Malaria Chemoprevention Campaigns in Senegal, Compared with the National Spontaneous Reporting System 1 2 3 4 • • • • Jean-Louis A. Ndiaye Ibrahima Diallo Youssoupha NDiaye Ekoue Kouevidjin 3 4 5 6 • • • • Ibrahima Aw Fassiatou Tairou Tidiane Ndoye Christine M. Halleux 4 4 2 4 • • • • Isaac Manga Mbaye Niang Dieme Medoune Ndiop Babacar Faye 6 6 4 7 • • • Piero Olliaro Corinne S. Merle Oumar Gaye Paul Milligan Published online: 1 June 2018 The Author(s) 2018 Abstract Methods Two strategies to improve reporting of AEs Background Seasonal malaria chemoprevention (SMC) during SMC campaigns were evaluated, in comparison using sulfadoxine–pyrimethamine plus amodiaquine has with the national system of spontaneous reporting, in 11 been introduced in 12 African countries. Additional health post areas in Senegal. In each health post, an average strategies for safety monitoring are needed to supplement of approximately 4000 children under 10 years of age national systems of spontaneous reporting that are known received SMC treatment each month for 3 months during to under represent the incidence of adverse reactions. the 2015 malaria transmission season—a total of 134,000 Objectives This study aimed to determine if adverse event treatments. In three health posts (serving approximately (AE) reporting could be improved using a smartphone 14,000 children), caregivers were encouraged to report any application provided to village health workers, or by active adverse reactions to the nurse at the health post or to a follow-up using a symptom card provided to caregivers. community health worker (CHW) in their village, who had been trained to use a smartphone application to report the Electronic supplementary material The online version of this event (enhanced spontaneous reporting). In two health article (https://doi.org/10.1007/s40290-018-0232-z) contains supple- posts (approximately 10,000 children), active follow-up of mentary material, which is available to authorized users. & Jean-Louis A. Ndiaye Medoune Ndiop jeanloab.ndiaye@ucad.edu.sn mnzop5@gmail.com Ibrahima Diallo Babacar Faye haril76@yahoo.fr bfaye67@yahoo.fr Youssoupha NDiaye Piero Olliaro youlebou@gmail.com olliarop@who.int Ekoue Kouevidjin Corinne S. Merle ekoue.kouevidjin@gmail.com merlec@who.int Ibrahima Aw Oumar Gaye ibouaw80@gmail.com oumar.gaye@ucad.edu.sn Fassiatou Tairou Paul Milligan fassiatht@yahoo.fr Paul.Milligan@lshtm.ac.uk Tidiane Ndoye Department of Parasitology, Cheikh Anta Diop University, tndoye16@gmail.com Thies University, Dakar, Senegal Christine M. Halleux National Malaria Control Programme, Ministry of Health and halleuxc@who.int Social Affairs, Dakar, Senegal Isaac Manga Sedhiou Medical Region, Ministry of Health and Social akhenmanga@yahoo.fr Affairs, Sedhiou, Senegal Mbaye Niang Dieme mbndieme@hotmail.com 190 J. A. Ndiaye et al. children at home was organized after each SMC campaign 1 Introduction to ask about AEs that caregivers had been asked to record on a symptom card (active surveillance). Six health posts Malaria remains a major public health concern in the (approximately 23,000 children) followed the national world, particularly in sub-Saharan Africa. The World system of spontaneous reporting using the national Health Organization (WHO) estimates 216 million malaria reporting (yellow) form. Each AE report was assessed by a cases and 445,000 malaria deaths occurred in 2016 [1], panel to determine likely association with SMC drugs. with the vast majority of malaria deaths occurring in Results The incidence of reported AEs was 2.4, 30.6, and children in sub-Saharan Africa caused by Plasmodium 21.6 per 1000 children treated per month, using the falciparum. Since 2012, the WHO has recommended sea- national system, enhanced spontaneous reporting, and sonal malaria chemoprevention (SMC), consisting of the active surveillance, respectively. The most commonly monthly administration of a full course of treatment with reported symptoms were vomiting, fever, and abdominal sulfadoxine–pyrimethamine (SP) and amodiaquine (AQ), pain. The incidence of vomiting, known to be caused by during the transmission season to prevent malaria. amodiaquine, was similar using both innovative methods Although SMC is recommended for children aged (10/1000 in the first month, decreasing to 2.5/1000 in the 3–59 months [2], in Senegal (as in many other areas where third month). Despite increased surveillance, no serious SMC is used) there is a substantial burden of severe adverse drug reactions were detected. malaria illness in older children. The Senegalese Ministry Conclusion Training CHWs in each village and health facility of Health made the decision to provide SMC for children staff to report AEs using a mobile phone application led to much up to 10 years of age when SMC was first introduced. After higher reporting rates than through the national system. This a pilot implementation in 2013, SMC was implemented in approach is feasible and acceptable, and could be further four regions of the country where malaria transmission was improved by strengthening laboratory investigation and the most intense (Kedougou, Kolda, Tambacounda and Sed- collection of control data immediately prior to SMC campaigns. hiou), in a population of approximately 600,000 children. From 2013 to 2017, approximately 8 million treatments were administered. A total of seven serious adverse events Key Points (SAEs) related to SMC have been reported in Senegal since the introduction of SMC, up to 2017: a case of Stevens– Seasonal malaria chemoprevention (SMC) is now Johnson syndrome and a case of toxic epidermal necroly- widely used to prevent malaria in children in West sis, both in 2014; a case of extrapyramidal syndrome and and Central Africa. Good safety monitoring is two cases of anaphylactic reactions in 2015; and one case essential to ensure SMC programs remain effective. of Stevens–Johnson syndrome and one anaphylactic reac- Training community health workers (CHWs) to tion in 2016. No SAEs were reported in 2017. Although recognize and report adverse events (AEs) improved pharmacovigilance during SMC campaigns was strength- detection of adverse drug reactions in this study ened through training of health staff to recognize adverse in Southern Senegal. reactions to SMC drugs, there is concern that adverse Training CHWs and health facility staff to report events (AEs) are being underreported. Known adverse using a mobile phone application enhanced safety reactions to SP and AQ have been reviewed by Phillips- reporting and improved timeliness of notifications Howard and Bjorkman [3] and, more recently, NDiaye during SMC campaigns. et al. [4]. AQ is associated with vomiting, extrapyramidal reactions, liver toxicity, and agranulocytosis, while SP can No serious AEs were detected despite enhanced cause liver toxicity and severe cutaneous reactions, surveillance. including Stevens–Johnson syndrome and toxic epidermal necrolysis, which are rare but life-threatening medical emergencies. Department of Parasitology, Cheikh Anta Diop University, Dakar, Senegal Pharmacovigilance systems based on spontaneous reporting are relatively simple and inexpensive to establish Department of Social Sciences, Cheikh Anta Diop University, Dakar, Senegal but often suffer from poor quality of reporting, as well as underreporting [5]. In addition, it is difficult to estimate The Special Programme for Research and Training in Tropical Diseases, World Health Organization, 1121 incidence rates of AEs through a spontaneous reporting Geneva 27, Switzerland system, generally because of lack of a denominator Department of Infectious Disease Epidemiology, London (number of people exposed). Additional methods are nee- School of Hygiene and Tropical Medicine, London ded to determine the safety profile of new medicines, as WC1E 7HT, UK Community-Based Safety Monitoring During Seasonal Malaria Chemoprevention Campaigns in Senegal 191 well as medicines put to new uses, in public health pro- administered are recorded on tally sheets that are collated grams [6]. This is especially important in the framework of to give monthly totals. mass administration of preventive treatment to children, delivered by community workers with limited medical 2.2 Study Design training. SAEs to SMC drugs appear to be uncommon, but, especially in the initial phases of SMC programs, there is a Eleven health posts in SMC implementation areas were need to supplement the national system of spontaneous selected and assigned to monitor SMC safety using one of reporting with additional methods of surveillance. three methods: safety monitoring employing the national The aim of this study was to evaluate whether AE system of spontaneous reporting using the national reporting could be improved using one of two approaches: reporting form (referred to as the national system), com- (1) reporting using a smartphone application when patients pleted by nurses or physicians at health facilities; reporting presented at the clinic or to a village health worker; and (2) using mobile phones (enhanced spontaneous reporting), active follow-up of children at home to ask about AEs and completed by nurses at health posts and by CHWs; and collect information caregivers were asked to record on a active follow-up of children at home after each SMC symptom card, compared with the current national system campaign to ask about AEs and collect information that for spontaneous reporting alone. Feasibility and accept- caregivers had been asked to record on a symptom card ability of the implementation of the strategies were also (active surveillance). Enhanced spontaneous reporting with assessed. mobile phones was implemented in two health posts in Kolda, and active surveillance was implemented in three health posts in Sedhiou. For comparison, three health posts 2 Methods in each district, where the staff were trained to report events using the national reporting form, acted as controls. 2.1 Study Population 2.3 Sample Size, Allocation, and Description This study was conducted in the health districts of Kolda of Surveillance Methods and Sedhiou in the south of Senegal (Electronic Supple- mentary Fig. S1) during the SMC campaign of 2015 (Au- In each district, two groups of health posts were selected, gust, September, and October). The population is primarily with a total of approximately 10,000 children per group. rural. In 2016, the under-5 mortality rate was estimated to The two groups were then randomly allocated within each be 100 per 1000 children, and the literacy rate was 55% in district to either control or enhanced or active surveillance. men and 41% in women [7]. The sample size of 10,000 children permits event rates Malaria is a major cause of severe illness in children. above 0.3/1000 children to be ruled out if zero events are Community case management for malaria is provided in observed, and is sufficient to give 99% probability of the more remote villages by a community health worker detecting at least one event if the rate is 1/2000 or more. As (Distributeur de Soins a` Domicile [DSDOM]) working SMC was administered to each child up to three times, the from their home or at a case de sante´ (health hut). SMC power for detecting events per 1000 treatments is greater. targeted children aged 3 months to 10 years and is deliv- Two health posts in Sedhiou (Diende ´ and Dembo Coly, ered door-to-door by the DSDOM and (in villages that do including 53 villages) were assigned to enhanced sponta- not have a resident health worker) by relais communau- neous reporting, and three health posts in Kolda (Ba- taires (community volunteers). In this report, DSDOM and gadadji, Dabo and Sikilo Ouest, including 106 villages) relais communautaires are collectively referred to as were assigned to active surveillance. Six health posts, three community health workers (CHWs). SMC delivery is in each district, assigned to the national system, acted as coordinated by the health post from where CHWs collect controls (Table 1). In the area assigned to active surveil- drugs each day, with SMC distribution following general lance, consent to participate in the study was sought when WHO recommendations [8]. One dose of SP and the first CHWs visited to administer the first monthly SMC dose of AQ were administered by the CHW on the first day, treatment. and the remaining two doses of AQ were left with the caregiver to administer over the next days. Children who 2.4 Enhanced Spontaneous Reporting are unwell are referred without treatment, however they may then receive SMC at the health post. CHWs are At the time of each monthly SMC distribution, caregivers trained to exclude children with a history of allergy to SMC were encouraged to contact the nearest CHW (DSDOM) or drugs and any child who had received SP, AQ, or a sulfa- the health post if the child was unwell after SMC admin- containing antibiotic in the previous month. Treatments istration. The nurses and the CHWs first entered the patient 192 J. A. Ndiaye et al. Table 1 Number of staff trained for pharmacovigilance in each health facility, size of the catchment population, number of children eligible for SMC, and number of children who received SMC treatment each month District Health post Method used for No. of health staff trained for Total SMC target No. of children who received PV PV: nurses (CHWs) population in population SMC treatment September October November Sedhiou Diende Enhanced 1 (9) 21,958 6587 4787 4849 5134 spontaneous reporting Dembo Enhanced 1 (9) 25,486 7646 7545 7574 7569 Coly spontaneous reporting Djibabouya National system 1 (10) 12,228 3668 2862 3008 3067 Bambaly National system 1 (10) 17,162 5148 3851 4100 4509 Djiredji National system 1 (10) 11,337 3401 3956 4004 3976 Kolda Bagadadji Active 1 (6) 14,862 4459 4393 4454 4576 surveillance Dabo Active 1 (6) 9705 2911 3057 3058 3065 surveillance Sikilo Active 1 (10) 8011 2403 2732 2994 3123 Ouest surveillance Sikolo Est National system 1 (10) 9412 2824 2794 2654 2675 Dioulacolon National system 1 (10) 16,619 4986 4997 4974 5121 Guiro Yero National system 1 (10) 9560 2868 2824 2925 2854 Boucar PV pharmacovigilance, CHWs community health workers, SMC seasonal malaria chemoprevention details in the consultation register and treated or referred 2.5 Active Surveillance the patient as appropriate. Then, in the case of children over 3 months and under 10 years of age who had received At the time of each monthly SMC administration, CHWs delivering SMC gave caregivers a card for each child, SMC and had any illness, the nurses and CHWs entered the name, age and sex of the child, all medicines received in which showed images illustrating fever, vomiting, rash, and pain. The caregiver was asked to tick the card if the the last month and the date of administration, and the date of onset of symptoms and a description of the symptoms, child had one of these symptoms or any other symptoms at into a smartphone application which then uploaded the data any time after SMC administration. The same CHWs went over the internet to a server in the project office in Dakar. back to each household after the end of each monthly The application was designed using Survey CTO (Dobility, campaign, between 6 and 12 days after the first day of the Inc., Wilmington, DE, USA), a platform based on Open SMC cycle, to ask caregivers about any AEs in children Data Kit (ODK), an open source set of programming tools who had received SMC. CHWs asked about additional for data capture. The date of the report was generated details, the date symptoms started, and, if the child no automatically. The phones were Samsung Duo, costing longer had symptoms, the date symptoms stopped, which the CHW then recorded on the card. The cards were col- CFA 30,000 each (US $51 at 2015 exchange rates), pro- vided with SIM cards and internet credit to cover the costs lected and taken to the health post, where the nurse reviewed them before entering the information into a of data uploading. The project team sent an SMS to each health worker each day to feed back to them the number of Microsoft Excel database. If the child was still unwell reports received from them the previous day and the total when the CHW visited, the child was referred to the nurse number from the start of the study. The team contacted at the health post, who completed a national reporting health workers who had not submitted any report the pre- (yellow) form. vious day to confirm there were no events. A database of all events reported was updated daily and shared with the 2.6 National System project team. If the nurse suspected the illness was an adverse drug reaction (ADR), a national reporting form The nurse at each health post was trained to complete a national reporting form (yellow form) for any child (yellow form) was completed. Community-Based Safety Monitoring During Seasonal Malaria Chemoprevention Campaigns in Senegal 193 presenting to the health post with suspected AEs, as per the symptoms tended to be reported together, odds ratios for national pharmacovigilance guidelines [6]. During SMC associations between symptom pairs were compared campaigns, yellow forms from all health posts were col- between groups, using a test of homogeneity, for the four lected at the district health center. District teams could then most commonly reported symptoms. Data were analyzed enter the forms into a Microsoft Excel database that was using Stata 13 (StataCorp LLC, College Station, TX, forwarded to the regional health team, or the forms were USA). sent to the regional team where they were entered. A final database was then sent to the Centre Anti-poison and the 2.10 Causality Assessment National Malaria Control Programme each month. Health workers were asked to report any illness as AEs not 2.7 Training necessarily having a causal relationship with medical treatment. The Centre Antipoison of Senegal and the An information sheet describing the symptoms of the pharmacovigilance Technical Committee analyzed the known AEs and highlighting vomiting, skin rash, and signs reports and assessed the severity and association with of jaundice was prepared and used to train health staff. In medicine intake following the WHO method [9]. All case all health posts, training for SMC delivery included health reports were submitted to the international drug monitoring facility staff and CHWs, and covered key messages about database through Vigiflow. pharmacovigilance, the symptoms of the known ADRs to SMC drugs, and how to report events using the national 2.11 Ethics form. In addition, for staff in health posts using the smartphones, training was held in each district over 2 days The research protocol and documents given to participants to explain how to report using the smartphone application. were submitted to the National Ethics Committee of Training included practical sessions on AE reporting, and Senegal and approval obtained prior to the start of the emphasized the importance of checking information before study. Administrative authorization was also given by the uploading, and the responsibilities of the nurses and CHWs Ministry of Health. Community consent was obtained for in pharmacovigilance. the phone-reporting pharmacovigilance prior to the begin- ning of the study. For children in the active surveillance 2.8 Sensitization and Study Preparation group, signed consent was sought from a parent or guardian after explaining the aims and procedures involved. An Meetings were held with regional and district medical advisory committee was set up by the National Malaria Control Programme at central level to provide guidance on officers to emphasize the importance of ensuring that lab- oratories had reagents for liver function tests and hema- the management of any severe AEs during SMC tology. In each district, a hospital pharmacist was campaigns. nominated to be the pharmacovigilance coordinator, and a study project manager was appointed. In all health posts, communities were informed about the SMC campaign, 3 Results including reminders to bring the child to a health worker if the child was sick after taking SMC medicine. In addition, A total of 1983 AEs were reported over the 3 months of in health posts with enhanced passive or active surveil- surveillance, out of a total of 134,061 monthly treatments. lance, additional community sensitization was organized Of these, 158 were reported through the national system (a by the health post nurse. rate of 2.4/1000 children treated/month), 1145 (31/1000 children/month) were reported through the enhanced 2.9 Data Management and Statistical Analysis spontaneous reporting system using CHWs and mobile phones, and 680 (22/1000 children/month) were reported Incidence rates were calculated per 1000 child months, through active surveillance. All patients reported having using the estimated number of children who received SMC taken SMC drugs, and 154/1983 (7.8%) had taken one or each month as the denominator. Rate ratios were used to more other medications in addition to SMC drugs, com- compare rates between surveillance methods, age groups, prising 53 who had taken camphorated tincture of opium and calendar months, estimated using Poisson regression (paregoric, a diarrhea treatment including 0.4 mg/mL with a robust standard error to account for clustering within morphine), 49 who had taken metopimazine (an antie- health posts and with stratification by district. The mean metic), 43 who had taken paracetamol, 24 who had taken number of symptoms reported per patient was compared zinc, 7 who had taken phloroglucinol (a treatment for with Poisson regression. To assess whether particular abdominal pain), 2 who had taken amoxicillin, 2 who had 194 J. A. Ndiaye et al. taken mequitazine (an antihistamine), 2 who had taken Rate per 1000 subjects per month metronidazole (an antibiotic used to treat protozoal infec- Naonal system tion), 1 who had taken chloroquine, and 1 who had taken Enhanced spontaneous reporng Acve surveillance traditional medicine. None of the events reported were considered serious. 3.1 Enhanced Spontaneous Reporting Overall, 1145 events were reported over 3 months, a rate of 30.6 [95% confidence interval (CI) 28.8–32.4] per 1000 children treated per month, compared with a rate of 1.65 (95% CI 1.27–2.15) per 1000 per month in health posts using the national system (Table 2). A total of 927 events (81% of the total) occurred within 10 days of the start of Fig. 1 Average incidence of each symptom, as reported through each the SMC cycle. The incidence of AEs decreased in each method of surveillance successive month, from 30.1/1000 in September, to 25.2 in October and 10.1 in November. The incidence was slightly SMC drugs, 27% of AEs were considered probably related lower in infants than in older children (Table 4). The most to SMC, 36% were possibly related, 1.2% were unlikely to commonly reported symptoms were fever, vomiting, and be related, and causality could not be assessed for the abdominal pain (Fig. 1). Among the older children remaining 36%. (5–10 years of age), the most common symptoms were, in In health posts in the same district that used the national descending order, fever, vomiting, abdominal pain, head- system of reporting, there were 55 events, a rate of 1.65/ ache, and diarrhea (Fig. 2), with a similar pattern being 1000/month. Rates similarly declined with each successive observed in the 12–59 months age group. Among children month. No AE was considered serious. When reports were aged 3–11 months, the most common symptoms were assessed for their association with SMC drugs, 11% were fever and diarrhea. Tiredness, cough, loss of appetite, considered probably related to SMC, 81% possibly related, dizziness, and pruritus were less commonly reported, and 1.2% unlikely to be related, and 6.3% could not be other symptoms (Electronic Supplementary Table S4) classified. together accounted for \ 1% of reported symptoms. No Enhanced reporting involving CHWs and using mobile AE was considered serious. The average number of phones increased reporting 18-fold (rate ratio 18.5, 95% CI symptoms reported per event was 1.67, with 50% of 8.65–39.7), and the incidence of episodes of vomiting by patients reporting more than one symptom. The distribution 10-fold (rate ratio 10.2, 95% CI 5.8–18.0). of the number of symptoms per event is shown in elec- tronic supplementary Table S2. Symptoms varied by age 3.2 Active Surveillance group (Fig. 2). Vomiting was reported at a rate of 10.7/ 1000/month (Table 3), and was most commonly associated Overall, 680 events were reported over 3 months, corre- with diarrhea or fever (electronic supplementary Table S3). sponding to a rate of 21.6 (95% CI 20.1–23.3) per 1000 When the reports were assessed for their relationship with children treated per month, compared with a rate of 3.24 Table 2 Incidence of adverse events (reports with one or more symptoms) following SMC distribution, using three methods of surveillance No. of events No. of treatments Incidence rate/1000 (95% CI) Incidence rate ratio (95% CI) Kolda National system 103 31,818 3.2 (2.7–3.9) Reference Active surveillance 680 31,452 21.6 (20.1–23.3) 6.7 (1.3–33.9) Sedhiou National system 55 33,333 1.65 (1.3–2.1) Reference Enhanced spontaneous reporting 1145 37,458 30.6 (28.8–32.4) 18.5 (8.65–39.7) SMC seasonal malaria chemoprevention, CI confidence interval Incidence rate ratios comparing the surveillance methods were estimated using Poisson regression, with the estimated number of SMC treatments as offset, and using robust standard errors to account for clustering within health post Community-Based Safety Monitoring During Seasonal Malaria Chemoprevention Campaigns in Senegal 195 Fig. 2 Incidence of adverse events, by age group, for each surveillance method. a Enhanced spontaneous reporting; b active surveillance Table 3 Incidence of vomiting following SMC distribution, using three methods of surveillance No. of events No. of treatments Incidence rate/1000 (95% CI) Incidence rate ratio (95% CI) Kolda National system 47 31,818 1.5 (1.1–2.0) Reference Active surveillance 335 31,452 10.7 (9.6–11.9) 7.2 (1.8–29.1) Sedhiou National system 34 33,333 1.0 (0.73–1.4) Reference Enhanced spontaneous reporting 389 37,458 10.4 (9.4–11.5) 10.2 (5.8–18.0) SMC seasonal malaria chemoprevention, CI confidence interval (95% CI 2.67–3.93) per 1000 per month in health posts percent of these reports were considered probably related using the national system. The incidence of AEs decreased to SMC, 70% were possibly related, while the remaining in each successive month, from 30.1/1000 in September, to 10% could not be classified. Active surveillance increased 25.2 in October and 10.1 in November. The incidence rate reporting almost sevenfold (rate ratio 6.7, 95% CI increased with increasing age, from 12.1/1000/month in 1.3–33.9) and reporting of episodes of vomiting by sev- infants, to 21.5 in children 12–59 months of age, and 24.0 enfold (rate ratio 7.2, 95% CI 1.8–29.1). in older children (Tables 4 and 5). The average number of symptoms reported per event was 1.2, with only one symptom reported in the majority (79%) of events. Among 4 Discussion the older children (5–10 years of age), the most common symptoms were, in descending order, vomiting, abdominal Although efforts have been made to strengthen national pain, fever, diarrhea, and headache. Vomiting was reported pharmacovigilance capacity during SMC programs [10], it at a rate of 10.7/1000/month. Other symptoms included is recognized that national systems based on spontaneous itching, lethargy, cough, loss of appetite, and dizziness, but reporting tend to underreport events. There might be these were uncommon. No AE was considered serious. multiple reasons for this: (1) events may not be recognized When reports were assessed for their association with SMC by the health worker as potentially related to administration drugs, 8.6% were considered probably related to SMC, of the medicine; (2) health staff may be too busy to report 72% possibly related, 0.24% were unlikely, and 19% could the event, or may not know how to report; or (3) the patient not be classified. may seek care outside the formal health system. Mild and In health posts in the same district which used the moderate events may be particularly underrepresented as national system of reporting, 103 events (3.24/1000/month) events may not be considered severe enough by a patient to were reported. No AE was considered serious. Twenty warrant consulting a health worker, or not considered by 196 J. A. Ndiaye et al. Table 4 Incidence of adverse No. of No. of treatments Incidence/1000 Incidence rate ratio (95% CI) events (reports with one or more events symptoms) following SMC distribution, using three Kolda methods of surveillance, by age Active surveillance group and month of Age group (months) administration 3–11 39 3212 12.1 0.56 (0.44–0.70) 12–59 324 15,052 21.5 Reference 60–120 317 13,188 24.0 1.12 (1.10–1.15) Month September 306 10,182 30.1 Reference October 265 10,506 25.2 0.84 (0.35–2.03) November 109 10,764 10.1 0.33 (0.09–1.26) Kolda National system Age group (months) 3–11 4 2842 1.4 0.45 (0.32–0.62) 12–59 48 15,255 3.1 Reference 60–120 51 13,721 3.7 1.16 (0.60–2.24) Month September 77 10,615 7.3 Reference October 19 10,553 1.8 0.25 (0.09–0.68) November 7 10,650 0.7 0.09 (0.03–0.25) Sedhiou Enhanced spontaneous reporting Age group (months) 3–11 114 3491 32.7 1.00 (0.73–1.38) 12–59 557 17,142 32.5 Reference 60–120 474 16,825 28.2 0.85 (0.77–0.94) Month September 570 12,332 46.2 Reference October 406 12,423 32.7 0.71 (0.38–1.30) November 169 12,703 13.3 0.29 (0.24–0.35) Sedhiou National system Age group (months) 3–11 8 2331 3.4 2.57 (1.89–3.48) 12–59 22 16,170 1.4 Reference 60–120 25 14,832 1.7 1.21 (0.46–3.19) Month September 31 10,669 2.9 Reference October 11 11,112 1.0 0.34 (0.24–0.48) November 13 11,552 1.1 0.38 (0.24–0.62) SMC seasonal malaria chemoprevention, CI confidence interval health workers to be important enough to report, but these considerably delayed. More intensive monitoring is there- could become a concern (jeopardizing acceptability or fore required, especially in the early phases of a new public adherence to treatment) if common. A further problem is health program, to establish its safety [6]. that reports are slow to reach national pharmacovigilance In October 2014, a workshop was organized by the and malaria program coordinators, therefore investigation Special Programme for Research and Training in Tropical (and the putting in place of any mitigating actions) may be Diseases (TDR) and the London School of Hygiene & Community-Based Safety Monitoring During Seasonal Malaria Chemoprevention Campaigns in Senegal 197 Table 5 Incidence of vomiting No. of events No. of treatments Rate/1000 (95% CI) Rate ratio (95% CI) following SMC distribution, using three methods of Kolda surveillance, by age group and Active surveillance month of administration Age group (months) 3–11 15 3212 4.67 (2.82–7.75) 0.47 (0.23–0.95) 12–59 147 15,052 9.77 (8.31–11.48) Reference 60–120 173 13,188 13.12 (11.30–15.23) 1.36 (1.10–1.67) Month September 171 10,182 16.79 (14.46–19.51) Reference October 117 10,506 11.14 (9.29–13.35) 0.66 (0.32–1.37) November 47 10,764 4.37 (3.28–5.81) 0.26 (0.07–0.99) Kolda National system Age group (months) 3–11 1 2842 0.35 (0.05–2.5) 0.28 (0.02,3.31) 12–59 19 15,255 1.2 (0.79–1.95) Reference 60–120 27 13,721 2.0 (1.3–2.9) 1.54 (0.51–4.66) Month September 41 10,615 3.9 (2.8–5.2) Reference October 5 10,553 0.47 (0.20–1.14) 0.12 (0.02–0.68) November 1 10,650 0.09 (0.01–0.67) 0.02 (0.00–0.13) Sedhiou Enhanced spontaneous reporting Age group (months) 3–11 15 3491 4.30 (2.59–7.13) 0.39 (0.21–0.70) 12–59 189 17,142 11.03 (9.56–12.72) Reference 60–120 185 16,825 11.00 (9.52–12.70) 0.98 (0.98–0.98) Month September 207 12,332 16.79 (14.65–19.24) Reference October 123 12,423 9.90 (8.30–11.82) 0.59 (0.42–0.82) November 59 12,703 4.65 (3.60–6.00) 0.28 (0.26–0.29) Sedhiou National system Age group (months) 3–11 4 2331 1.72 (0.64–4.57) 2.82 (1.10–7.28) 12–59 10 16,170 0.62 (0.33–1.15) Reference 60–120 20 14,832 1.35 (0.87–2.09) 2.13 (0.89–5.10) Month September 20 10,669 1.88 (1.21–2.91) Reference October 6 11,112 0.54 (0.24–1.20) 0.29 (0.11–0.77) November 8 11,552 0.69 (0.35–1.39) 0.37 (0.15–0.94) SMC seasonal malaria chemoprevention, CI confidence interval Tropical Medicine (LSHTM) to bring together coordinators workshop highlighted the need for operational research of national malaria control programs and national phar- into innovative practical methods that could simplify and macovigilance centers from countries that had introduced expedite reporting and improve promptness of notification SMC, or planned to in the near future, to discuss with of events to the national malaria program and the phar- pharmacovigilance experts approaches that could be used macovigilance center [11]. This led to planning of the to improve safety monitoring of SMC programs. The present study that evaluated two approaches to community- 198 J. A. Ndiaye et al. based pharmacovigilance involving CHWs; one based on would remain, as, in this population, morbidity is seasonal, passive, spontaneous reporting in which CHWs and health with the rainy season being associated with the transmis- facility staff were trained to report AEs using a mobile sion of infections and with malnutrition. In this study, in phone application, and the other approach involving active the enhanced and active surveillance areas, health workers follow-up by CHWs after each SMC campaign to ask about were encouraged to report all events. When reports were AEs that caregivers had been asked to record on a symptom assessed for evidence of a relationship with SMC drugs, card. Both methods involved sensitization of the commu- over one-third could not be reliably classified. Therefore, nity to the importance of reporting suspected AEs, and a the overall rates are likely to overestimate the true rates of central supervision team to process, share, and analyze AEs to SMC drugs, but the rates (or rather the upper reports and to provide feedback to reporters. confidence limits on the rates) can be interpreted as rep- The incidence rate of reported events was higher using resenting an upper bound for the true rate of drug-related community-based spontaneous reporting—31/1000 chil- symptoms. Furthermore, there may have been reporting dren/month, compared with 22/1000 children/month using bias, both by caregivers who may have been less likely to active surveillance. These rates compare with rates of report symptoms not listed on the symptom card, and by 1.7–3.2 through the national system in the same districts. CHWs, whose training emphasized the known side effects The greater incidence in the enhanced spontaneous of SMC drugs. A further limitation of the study (in com- reporting arm of the study reflects the longer period of time mon with routine pharmacovigilance) is that although steps over which events were observed. In the active surveillance were taken to improve laboratory capacity for routine arm of the study, participants were asked about events investigation, it was not possible to routinely perform occurring up to 10 days prior to the visit, whereas any hematological analysis or liver function tests, and therefore event during the month could be reported in the enhanced it is possible that cases of clinically silent neutropenia spontaneous reporting arm. (which can be associated with AQ) and hepatitis (which When specific symptoms were considered, the incidence can be associated with both AQ and SP) may have gone rate of vomiting, which, it is known, can be caused by AQ, undetected. Severe cutaneous reactions, known to be was seen to be similar in the active and enhanced sponta- associated with SP, and extrapyramidal syndrome, known neous reporting arms of the study in infants and young to be caused by AQ, are severe and unmistakable, and the children, and slightly more common among older children fact that no cases of these syndromes were seen suggests it in the active arm of the study, while fever and diarrhea is unlikely that any such events occurred during the study were more commonly reported in the spontaneous report- period in the populations under enhanced or active ing arm of the study. It was hypothesized before the study surveillance. The most commonly reported symptoms were gas- that mild side effects would not be reported spontaneously and active surveillance would be necessary to determine trointestinal disorders (nausea/vomiting, abdominal pain, the true burden. The fact that the incidence of vomiting was diarrhea), fever, headache, tiredness, cough, loss of appe- similar using both methods suggests this is not the case tite, dizziness and pruritus. These symptoms were similar when good access is provided through CHWs based in the to those reported in other studies of antimalarials in chil- village. In the enhanced spontaneous reporting arm of the dren in the Central African Republic [13], Burkina Faso study, 16 CHWs were based in the community, to whom [14], and Senegal [4], and in school children in Uganda patients could report, in addition to the two health post [15]. Fever was more commonly reported in our study nurses. It is likely that this improved access for patients, compared with other studies on SP ? AQ, but this may and the provision of training for these staff to recognize reflect that the definition of fever includes a reported his- AEs and when and how to report, may have contributed to tory of fever, and that the study took place over an the high reporting rate, as has been found in other studies extended period during the rainy season, when febrile ill- [12]. nesses are common. None of the reported AEs was classified as serious. As The incidence of AEs decreased in successive rounds of the number of children who received SMC in the intensi- SMC, a finding that is consistent with the results of others fied surveillance area was at least 23,000, the upper 95% studies [4, 14]. The fact that this effect was observed in confidence limit for the rate of severe AEs per child was 1 active and passive surveillance areas suggests there may in 7800 (i.e. on the basis of this study we can rule out rates have been gradual tolerization to the effects of AQ; greater than this). mothers could also become less likely to report mild A limitation of this study is the lack of suitable controls symptoms over time. to establish the rate of symptoms in children who did not For reports submitted by mobile phone, the average delay receive SMC. Data could have been collected prior to the from case presentation to the notification reaching the cen- start of the SMC campaigns, although some confounding tral office was 24 h. In a study using SMS to monitor AEs Community-Based Safety Monitoring During Seasonal Malaria Chemoprevention Campaigns in Senegal 199 Funding This study was funded by the Special Programme for after immunization in Australia, most patients who received Research and Training in Tropical Diseases. Jean-Louis A. Ndiaye is an SMS query about adverse vaccine reactions sent a supported by a Wellcome Trust Intermediate Fellowship in Public response within 2 h of receiving the query [16]. A study in Health and Tropical Medicine which also funded the open access fee Ghana, where the use of mobile phones to contact patients for this article. was compared with home visits to identify AEs related to Conflict of interest Jean-Louis A. Ndiaye, Ibrahima Diallo, Yous- artemisinin-based combination therapy for treating uncom- soupha NDiaye, Ekoue Kouevidjin, Ibrahima Aw, Fassiatou Tairou, plicated malaria found a slightly higher reporting rate with Tidiane Ndoye, Christine M. Halleux, Isaac Manga, Mbaye Niang mobile phones, and that it was possible to interview the Dieme, Medoune Ndiop, Babacar Faye, Piero Olliaro, Corinne S. Merle, Oumar Gaye, and Paul Milligan declare that they have no majority of patients within a few days [17]. In Uganda, an conflicts of interest. Piero Olliaro, Corinne S. Merle, and Christine M. SMS-based reporting system for monitoring malaria diag- Halleux are staff members of the WHO. The views expressed in this nosis and treatment improved timeliness in data reporting publication are the views of the authors alone and do not necessarily [18]. A pilot study in rural districts of Kenya using mobile represent the decisions, policies, or views of the WHO. text messaging for malaria surveillance demonstrated that Open Access This article is distributed under the terms of the Creative phone reporting can improve timeliness and reporting of Commons Attribution 3.0 IGO License (http://creativecommons.org/ data [19]. Semi-structured interviews and focus group dis- licenses/by/3.0/igo/), which permits unrestricted use, duplication, cussions conducted in conjunction with this study indicated adaptation, distribution, and reproduction in any medium or format, as that active surveillance was appreciated by community long as you give appropriate credit to the original author(s) and the source. In any reproduction of this article there should not be any members, but CHWs found this time-consuming and con- suggestion that WHO or this article endorse any specific organization sidered enhanced spontaneous reporting to be effective; or products. The use of the WHO logo is not permitted. This notice however, technical difficulties in terms of electricity supply should be preserved along with the article’s original URL. and internet connections were noted. The authors intend to publish results from this aspect of the study separately. References 1. World Health Organization. World Malaria Report 2016. http:// www.who.int/malaria/publications/world-malaria-report-2017/ 5 Conclusions report/en/. Accessed 14 Dec 2017. 2. World Health Organization. WHO Policy Recommendation: Sea- This study has shown that involving CHWs in safety sonal Malaria Chemoprevention (SMC) for Plasmodium falci- reporting, and training nurses and CHWs to report using a parum malaria control in highly seasonal transmission areas of the Sahel sub-region in Africa. March 2012 (in English and French). mobile phone application, can be used to enhance safety 2012. http://www.who.int/malaria/publications/atoz/who_smc_ reporting and improve timeliness of notifications, but the policy_recommendation/en/index.html. Accessed 15 Apr 2016. approach relies on training and supervision of CHWs and 3. Phillips-Howard PA, Bjorkman AB. Ascertainment of risk of health facility staff, effective community sensitization, and serious adverse reactions associated with chemoprophylactic antimalarial drugs. Bull World Health Organ. 1990;68(4):493–504. a central team to process reports and provide feedback. 4. NDiaye JL, Cisse B, Ba EH, Gomis JF, Ndour CT, Molez JF, These strategies could be adopted in sentinel sites when et al. Safety of Seasonal Malaria Chemoprevention (SMC) with new public health programs are introduced and scaled up. sulfadoxine–pyrimethamine plus amodiaquine when delivered to Enhanced safety monitoring using this approach could be children under 10 years of age by district health services in Senegal: results from a stepped wedge cluster randomized trial. established in SMC areas; however, for comparison, lab- PLoS One. 2016;11(12):e0168421. oratory investigation should be strengthened and baseline 5. Pal SN, Duncombe C, Falzon D, Olsson S. WHO strategy for incidence should be collected prior to SMC campaigns. collecting safety data in public health programmes: complementing spontaneous reporting systems. Drug Saf. 2013;36(2):75–81. Acknowledgements The authors are grateful to the district medical 6. World Health Organization. The safety of medicine in public teams of Kolda and Sedhiou; the regional medical teams; the com- health programmes: pharmacovigilance an essential tool. Geneva: munities and health staff in the study areas; the National Malaria World Health Organization; 2016. Control Programme, Senegal; the staff of the Centre Anti-Poison, 7. ANSD. Senegal: Enquete Demographique et de Sante Continue Dakar; the Direction de la Pharmacie et des Medicaments, Dakar; and (EDS-Continue) 2016. Rockville, MD. Agence Nationale de la the Faculty of Medecine, University Cheikh Anta Diop. Statistique ertde la Demographie, Daka, Senegal nd the DHS Program, ICF; 2017. Compliance with Ethical Standards 8. World Health Organization. Seasonal malaria chemoprevention with sulfadoxine-pyrimethamine plus amodiaquine in children: a field guide. Geneva: World Health Organization; 2013. http:// Ethical approval All procedures performed involving human par- www.who.int/malaria/publications/atoz/9789241504737/en/. Acces- ticipants were in accordance with the ethical standards of the Ethics Committee in Senegal and with the 1964 Helsinki declaration and its sed 15 Dec 2017. later amendments or comparable ethical standards. Informed consent 9. World Health Organization. The use of the WHO-UMC system was sought from all participants by explaining the aims and activities for standardised case causality assessment. 2012. http://www. prior to seeking signed consent. who-umc.org/Graphics/26649.pdf. Accessed 27 July 2016. 200 J. A. Ndiaye et al. 10. World Health Organization. Integrating pharmacovigilance in dihydroartemisinin–piperaquine compared with sulfadoxine-pyr- Seasonal Malaria Chemoprevention: the story so far. Pharma- imethamine plus amodiaquine for seasonal malaria chemopre- ceuticals Newsletter. 2017;4:33–34. http://apps.who.int/ vention in Burkina Faso. Antimicrob Agents Chemother. medicinedocs/documents/s23296en/s23296en.pdf. Accessed 15 2015;59(8):4387–96. Dec 2017. 15. Nankabirwa J, Cundill B, Clarke S, Kabatereine N, Rosenthal PJ, 11. World Health Organization Special Programme for Research and Dorsey G, et al. Efficacy, safety, and tolerability of three regi- Training in Tropical Diseases (TDR). Strengthening safety mens for prevention of malaria: a randomized, placebo-controlled surveillance in seasonal malaria chemoprevention campaigns in trial in Ugandan schoolchildren. PLOS One. 2010;5(10):e13438. Africa. 2015. http://www.who.int/tdr/news/2015/safety_surv_ 16. Leeb A, Regan AK, Peters IJ, Leeb C, Leeb G, Effler PV. Using mal_chemoprev_camp/en/. Accessed 15 Dec 2017. automated text messages to monitor adverse events following 12. Kabanywanyi AM, Mulure N, Migoha C, Malila A, Lengeler C, immunisation in general practice. Med J Aust. Schlienger R. Experience of safety monitoring in the context of a 2014;200(7):416–8. prospective observational study of artemether–lumefantrine in 17. Kukula VA, Dodoo AA, Akpakli J, Narh-Bana SA, Clerk C, rural Tanzania: lessons learned for pharmacovigilance reporting. Adjei A, et al. Feasibility and cost of using mobile phones for Malar J. 2010;9:205. capturing drug safety information in peri-urban settlement in 13. Djalle ´ D, Njuimo SP, Manirakiza A, Laganier R, Le Faou A, Ghana: a prospective cohort study of patients with uncomplicated Rogier C. Efficacy and safety of artemether ? lumefantrine, malaria. Malar J. 2015;14(1):411. artesunate ? sulphamethoxypyrazine-pyrimethamine and arte- 18. Asiimwe C, Gelvin D, Lee E, Ben Amor Y, Quinto E, Katureebe sunate ? amodiaquine and sulphadoxine- C, et al. Use of an innovative, affordable, and open-source short pyrimethamine ? amodiaquine in the treatment of uncompli- message service-based tool to monitor malaria in remote areas of cated falciparum malaria in Bangui, Central African Republic: a Uganda. Am J Trop Med Hyg. 2011;85(1):26–33. randomized trial. Malar J. 2014;13:9. 19. Githinji S, Kigen S, Memusi D, Nyandigisi A, Wamari A, Muturi 14. Zongo I, Milligan P, Compaore YD, Some AF, Greenwood B, A. Using mobile phone text messaging for malaria surveillance in Tarning J, et al. Randomized noninferiority trial of rural Kenya. Malar J. 2014;13:107. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Pharmaceutical Medicine Springer Journals

Evaluation of Two Strategies for Community-Based Safety Monitoring during Seasonal Malaria Chemoprevention Campaigns in Senegal, Compared with the National Spontaneous Reporting System

Free
12 pages

Loading next page...
 
/lp/springer_journal/evaluation-of-two-strategies-for-community-based-safety-monitoring-QTxuZsNj0k
Publisher
Springer International Publishing
Copyright
Copyright © 2018 by The Author(s)
Subject
Biomedicine; Pharmaceutical Sciences/Technology; Pharmacotherapy; Pharmacology/Toxicology
ISSN
1178-2595
eISSN
1179-1993
D.O.I.
10.1007/s40290-018-0232-z
Publisher site
See Article on Publisher Site

Abstract

Pharm Med (2018) 32:189–200 https://doi.org/10.1007/s40290-018-0232-z OR IGINAL RESEARCH ARTIC L E Evaluation of Two Strategies for Community-Based Safety Monitoring during Seasonal Malaria Chemoprevention Campaigns in Senegal, Compared with the National Spontaneous Reporting System 1 2 3 4 • • • • Jean-Louis A. Ndiaye Ibrahima Diallo Youssoupha NDiaye Ekoue Kouevidjin 3 4 5 6 • • • • Ibrahima Aw Fassiatou Tairou Tidiane Ndoye Christine M. Halleux 4 4 2 4 • • • • Isaac Manga Mbaye Niang Dieme Medoune Ndiop Babacar Faye 6 6 4 7 • • • Piero Olliaro Corinne S. Merle Oumar Gaye Paul Milligan Published online: 1 June 2018 The Author(s) 2018 Abstract Methods Two strategies to improve reporting of AEs Background Seasonal malaria chemoprevention (SMC) during SMC campaigns were evaluated, in comparison using sulfadoxine–pyrimethamine plus amodiaquine has with the national system of spontaneous reporting, in 11 been introduced in 12 African countries. Additional health post areas in Senegal. In each health post, an average strategies for safety monitoring are needed to supplement of approximately 4000 children under 10 years of age national systems of spontaneous reporting that are known received SMC treatment each month for 3 months during to under represent the incidence of adverse reactions. the 2015 malaria transmission season—a total of 134,000 Objectives This study aimed to determine if adverse event treatments. In three health posts (serving approximately (AE) reporting could be improved using a smartphone 14,000 children), caregivers were encouraged to report any application provided to village health workers, or by active adverse reactions to the nurse at the health post or to a follow-up using a symptom card provided to caregivers. community health worker (CHW) in their village, who had been trained to use a smartphone application to report the Electronic supplementary material The online version of this event (enhanced spontaneous reporting). In two health article (https://doi.org/10.1007/s40290-018-0232-z) contains supple- posts (approximately 10,000 children), active follow-up of mentary material, which is available to authorized users. & Jean-Louis A. Ndiaye Medoune Ndiop jeanloab.ndiaye@ucad.edu.sn mnzop5@gmail.com Ibrahima Diallo Babacar Faye haril76@yahoo.fr bfaye67@yahoo.fr Youssoupha NDiaye Piero Olliaro youlebou@gmail.com olliarop@who.int Ekoue Kouevidjin Corinne S. Merle ekoue.kouevidjin@gmail.com merlec@who.int Ibrahima Aw Oumar Gaye ibouaw80@gmail.com oumar.gaye@ucad.edu.sn Fassiatou Tairou Paul Milligan fassiatht@yahoo.fr Paul.Milligan@lshtm.ac.uk Tidiane Ndoye Department of Parasitology, Cheikh Anta Diop University, tndoye16@gmail.com Thies University, Dakar, Senegal Christine M. Halleux National Malaria Control Programme, Ministry of Health and halleuxc@who.int Social Affairs, Dakar, Senegal Isaac Manga Sedhiou Medical Region, Ministry of Health and Social akhenmanga@yahoo.fr Affairs, Sedhiou, Senegal Mbaye Niang Dieme mbndieme@hotmail.com 190 J. A. Ndiaye et al. children at home was organized after each SMC campaign 1 Introduction to ask about AEs that caregivers had been asked to record on a symptom card (active surveillance). Six health posts Malaria remains a major public health concern in the (approximately 23,000 children) followed the national world, particularly in sub-Saharan Africa. The World system of spontaneous reporting using the national Health Organization (WHO) estimates 216 million malaria reporting (yellow) form. Each AE report was assessed by a cases and 445,000 malaria deaths occurred in 2016 [1], panel to determine likely association with SMC drugs. with the vast majority of malaria deaths occurring in Results The incidence of reported AEs was 2.4, 30.6, and children in sub-Saharan Africa caused by Plasmodium 21.6 per 1000 children treated per month, using the falciparum. Since 2012, the WHO has recommended sea- national system, enhanced spontaneous reporting, and sonal malaria chemoprevention (SMC), consisting of the active surveillance, respectively. The most commonly monthly administration of a full course of treatment with reported symptoms were vomiting, fever, and abdominal sulfadoxine–pyrimethamine (SP) and amodiaquine (AQ), pain. The incidence of vomiting, known to be caused by during the transmission season to prevent malaria. amodiaquine, was similar using both innovative methods Although SMC is recommended for children aged (10/1000 in the first month, decreasing to 2.5/1000 in the 3–59 months [2], in Senegal (as in many other areas where third month). Despite increased surveillance, no serious SMC is used) there is a substantial burden of severe adverse drug reactions were detected. malaria illness in older children. The Senegalese Ministry Conclusion Training CHWs in each village and health facility of Health made the decision to provide SMC for children staff to report AEs using a mobile phone application led to much up to 10 years of age when SMC was first introduced. After higher reporting rates than through the national system. This a pilot implementation in 2013, SMC was implemented in approach is feasible and acceptable, and could be further four regions of the country where malaria transmission was improved by strengthening laboratory investigation and the most intense (Kedougou, Kolda, Tambacounda and Sed- collection of control data immediately prior to SMC campaigns. hiou), in a population of approximately 600,000 children. From 2013 to 2017, approximately 8 million treatments were administered. A total of seven serious adverse events Key Points (SAEs) related to SMC have been reported in Senegal since the introduction of SMC, up to 2017: a case of Stevens– Seasonal malaria chemoprevention (SMC) is now Johnson syndrome and a case of toxic epidermal necroly- widely used to prevent malaria in children in West sis, both in 2014; a case of extrapyramidal syndrome and and Central Africa. Good safety monitoring is two cases of anaphylactic reactions in 2015; and one case essential to ensure SMC programs remain effective. of Stevens–Johnson syndrome and one anaphylactic reac- Training community health workers (CHWs) to tion in 2016. No SAEs were reported in 2017. Although recognize and report adverse events (AEs) improved pharmacovigilance during SMC campaigns was strength- detection of adverse drug reactions in this study ened through training of health staff to recognize adverse in Southern Senegal. reactions to SMC drugs, there is concern that adverse Training CHWs and health facility staff to report events (AEs) are being underreported. Known adverse using a mobile phone application enhanced safety reactions to SP and AQ have been reviewed by Phillips- reporting and improved timeliness of notifications Howard and Bjorkman [3] and, more recently, NDiaye during SMC campaigns. et al. [4]. AQ is associated with vomiting, extrapyramidal reactions, liver toxicity, and agranulocytosis, while SP can No serious AEs were detected despite enhanced cause liver toxicity and severe cutaneous reactions, surveillance. including Stevens–Johnson syndrome and toxic epidermal necrolysis, which are rare but life-threatening medical emergencies. Department of Parasitology, Cheikh Anta Diop University, Dakar, Senegal Pharmacovigilance systems based on spontaneous reporting are relatively simple and inexpensive to establish Department of Social Sciences, Cheikh Anta Diop University, Dakar, Senegal but often suffer from poor quality of reporting, as well as underreporting [5]. In addition, it is difficult to estimate The Special Programme for Research and Training in Tropical Diseases, World Health Organization, 1121 incidence rates of AEs through a spontaneous reporting Geneva 27, Switzerland system, generally because of lack of a denominator Department of Infectious Disease Epidemiology, London (number of people exposed). Additional methods are nee- School of Hygiene and Tropical Medicine, London ded to determine the safety profile of new medicines, as WC1E 7HT, UK Community-Based Safety Monitoring During Seasonal Malaria Chemoprevention Campaigns in Senegal 191 well as medicines put to new uses, in public health pro- administered are recorded on tally sheets that are collated grams [6]. This is especially important in the framework of to give monthly totals. mass administration of preventive treatment to children, delivered by community workers with limited medical 2.2 Study Design training. SAEs to SMC drugs appear to be uncommon, but, especially in the initial phases of SMC programs, there is a Eleven health posts in SMC implementation areas were need to supplement the national system of spontaneous selected and assigned to monitor SMC safety using one of reporting with additional methods of surveillance. three methods: safety monitoring employing the national The aim of this study was to evaluate whether AE system of spontaneous reporting using the national reporting could be improved using one of two approaches: reporting form (referred to as the national system), com- (1) reporting using a smartphone application when patients pleted by nurses or physicians at health facilities; reporting presented at the clinic or to a village health worker; and (2) using mobile phones (enhanced spontaneous reporting), active follow-up of children at home to ask about AEs and completed by nurses at health posts and by CHWs; and collect information caregivers were asked to record on a active follow-up of children at home after each SMC symptom card, compared with the current national system campaign to ask about AEs and collect information that for spontaneous reporting alone. Feasibility and accept- caregivers had been asked to record on a symptom card ability of the implementation of the strategies were also (active surveillance). Enhanced spontaneous reporting with assessed. mobile phones was implemented in two health posts in Kolda, and active surveillance was implemented in three health posts in Sedhiou. For comparison, three health posts 2 Methods in each district, where the staff were trained to report events using the national reporting form, acted as controls. 2.1 Study Population 2.3 Sample Size, Allocation, and Description This study was conducted in the health districts of Kolda of Surveillance Methods and Sedhiou in the south of Senegal (Electronic Supple- mentary Fig. S1) during the SMC campaign of 2015 (Au- In each district, two groups of health posts were selected, gust, September, and October). The population is primarily with a total of approximately 10,000 children per group. rural. In 2016, the under-5 mortality rate was estimated to The two groups were then randomly allocated within each be 100 per 1000 children, and the literacy rate was 55% in district to either control or enhanced or active surveillance. men and 41% in women [7]. The sample size of 10,000 children permits event rates Malaria is a major cause of severe illness in children. above 0.3/1000 children to be ruled out if zero events are Community case management for malaria is provided in observed, and is sufficient to give 99% probability of the more remote villages by a community health worker detecting at least one event if the rate is 1/2000 or more. As (Distributeur de Soins a` Domicile [DSDOM]) working SMC was administered to each child up to three times, the from their home or at a case de sante´ (health hut). SMC power for detecting events per 1000 treatments is greater. targeted children aged 3 months to 10 years and is deliv- Two health posts in Sedhiou (Diende ´ and Dembo Coly, ered door-to-door by the DSDOM and (in villages that do including 53 villages) were assigned to enhanced sponta- not have a resident health worker) by relais communau- neous reporting, and three health posts in Kolda (Ba- taires (community volunteers). In this report, DSDOM and gadadji, Dabo and Sikilo Ouest, including 106 villages) relais communautaires are collectively referred to as were assigned to active surveillance. Six health posts, three community health workers (CHWs). SMC delivery is in each district, assigned to the national system, acted as coordinated by the health post from where CHWs collect controls (Table 1). In the area assigned to active surveil- drugs each day, with SMC distribution following general lance, consent to participate in the study was sought when WHO recommendations [8]. One dose of SP and the first CHWs visited to administer the first monthly SMC dose of AQ were administered by the CHW on the first day, treatment. and the remaining two doses of AQ were left with the caregiver to administer over the next days. Children who 2.4 Enhanced Spontaneous Reporting are unwell are referred without treatment, however they may then receive SMC at the health post. CHWs are At the time of each monthly SMC distribution, caregivers trained to exclude children with a history of allergy to SMC were encouraged to contact the nearest CHW (DSDOM) or drugs and any child who had received SP, AQ, or a sulfa- the health post if the child was unwell after SMC admin- containing antibiotic in the previous month. Treatments istration. The nurses and the CHWs first entered the patient 192 J. A. Ndiaye et al. Table 1 Number of staff trained for pharmacovigilance in each health facility, size of the catchment population, number of children eligible for SMC, and number of children who received SMC treatment each month District Health post Method used for No. of health staff trained for Total SMC target No. of children who received PV PV: nurses (CHWs) population in population SMC treatment September October November Sedhiou Diende Enhanced 1 (9) 21,958 6587 4787 4849 5134 spontaneous reporting Dembo Enhanced 1 (9) 25,486 7646 7545 7574 7569 Coly spontaneous reporting Djibabouya National system 1 (10) 12,228 3668 2862 3008 3067 Bambaly National system 1 (10) 17,162 5148 3851 4100 4509 Djiredji National system 1 (10) 11,337 3401 3956 4004 3976 Kolda Bagadadji Active 1 (6) 14,862 4459 4393 4454 4576 surveillance Dabo Active 1 (6) 9705 2911 3057 3058 3065 surveillance Sikilo Active 1 (10) 8011 2403 2732 2994 3123 Ouest surveillance Sikolo Est National system 1 (10) 9412 2824 2794 2654 2675 Dioulacolon National system 1 (10) 16,619 4986 4997 4974 5121 Guiro Yero National system 1 (10) 9560 2868 2824 2925 2854 Boucar PV pharmacovigilance, CHWs community health workers, SMC seasonal malaria chemoprevention details in the consultation register and treated or referred 2.5 Active Surveillance the patient as appropriate. Then, in the case of children over 3 months and under 10 years of age who had received At the time of each monthly SMC administration, CHWs delivering SMC gave caregivers a card for each child, SMC and had any illness, the nurses and CHWs entered the name, age and sex of the child, all medicines received in which showed images illustrating fever, vomiting, rash, and pain. The caregiver was asked to tick the card if the the last month and the date of administration, and the date of onset of symptoms and a description of the symptoms, child had one of these symptoms or any other symptoms at into a smartphone application which then uploaded the data any time after SMC administration. The same CHWs went over the internet to a server in the project office in Dakar. back to each household after the end of each monthly The application was designed using Survey CTO (Dobility, campaign, between 6 and 12 days after the first day of the Inc., Wilmington, DE, USA), a platform based on Open SMC cycle, to ask caregivers about any AEs in children Data Kit (ODK), an open source set of programming tools who had received SMC. CHWs asked about additional for data capture. The date of the report was generated details, the date symptoms started, and, if the child no automatically. The phones were Samsung Duo, costing longer had symptoms, the date symptoms stopped, which the CHW then recorded on the card. The cards were col- CFA 30,000 each (US $51 at 2015 exchange rates), pro- vided with SIM cards and internet credit to cover the costs lected and taken to the health post, where the nurse reviewed them before entering the information into a of data uploading. The project team sent an SMS to each health worker each day to feed back to them the number of Microsoft Excel database. If the child was still unwell reports received from them the previous day and the total when the CHW visited, the child was referred to the nurse number from the start of the study. The team contacted at the health post, who completed a national reporting health workers who had not submitted any report the pre- (yellow) form. vious day to confirm there were no events. A database of all events reported was updated daily and shared with the 2.6 National System project team. If the nurse suspected the illness was an adverse drug reaction (ADR), a national reporting form The nurse at each health post was trained to complete a national reporting form (yellow form) for any child (yellow form) was completed. Community-Based Safety Monitoring During Seasonal Malaria Chemoprevention Campaigns in Senegal 193 presenting to the health post with suspected AEs, as per the symptoms tended to be reported together, odds ratios for national pharmacovigilance guidelines [6]. During SMC associations between symptom pairs were compared campaigns, yellow forms from all health posts were col- between groups, using a test of homogeneity, for the four lected at the district health center. District teams could then most commonly reported symptoms. Data were analyzed enter the forms into a Microsoft Excel database that was using Stata 13 (StataCorp LLC, College Station, TX, forwarded to the regional health team, or the forms were USA). sent to the regional team where they were entered. A final database was then sent to the Centre Anti-poison and the 2.10 Causality Assessment National Malaria Control Programme each month. Health workers were asked to report any illness as AEs not 2.7 Training necessarily having a causal relationship with medical treatment. The Centre Antipoison of Senegal and the An information sheet describing the symptoms of the pharmacovigilance Technical Committee analyzed the known AEs and highlighting vomiting, skin rash, and signs reports and assessed the severity and association with of jaundice was prepared and used to train health staff. In medicine intake following the WHO method [9]. All case all health posts, training for SMC delivery included health reports were submitted to the international drug monitoring facility staff and CHWs, and covered key messages about database through Vigiflow. pharmacovigilance, the symptoms of the known ADRs to SMC drugs, and how to report events using the national 2.11 Ethics form. In addition, for staff in health posts using the smartphones, training was held in each district over 2 days The research protocol and documents given to participants to explain how to report using the smartphone application. were submitted to the National Ethics Committee of Training included practical sessions on AE reporting, and Senegal and approval obtained prior to the start of the emphasized the importance of checking information before study. Administrative authorization was also given by the uploading, and the responsibilities of the nurses and CHWs Ministry of Health. Community consent was obtained for in pharmacovigilance. the phone-reporting pharmacovigilance prior to the begin- ning of the study. For children in the active surveillance 2.8 Sensitization and Study Preparation group, signed consent was sought from a parent or guardian after explaining the aims and procedures involved. An Meetings were held with regional and district medical advisory committee was set up by the National Malaria Control Programme at central level to provide guidance on officers to emphasize the importance of ensuring that lab- oratories had reagents for liver function tests and hema- the management of any severe AEs during SMC tology. In each district, a hospital pharmacist was campaigns. nominated to be the pharmacovigilance coordinator, and a study project manager was appointed. In all health posts, communities were informed about the SMC campaign, 3 Results including reminders to bring the child to a health worker if the child was sick after taking SMC medicine. In addition, A total of 1983 AEs were reported over the 3 months of in health posts with enhanced passive or active surveil- surveillance, out of a total of 134,061 monthly treatments. lance, additional community sensitization was organized Of these, 158 were reported through the national system (a by the health post nurse. rate of 2.4/1000 children treated/month), 1145 (31/1000 children/month) were reported through the enhanced 2.9 Data Management and Statistical Analysis spontaneous reporting system using CHWs and mobile phones, and 680 (22/1000 children/month) were reported Incidence rates were calculated per 1000 child months, through active surveillance. All patients reported having using the estimated number of children who received SMC taken SMC drugs, and 154/1983 (7.8%) had taken one or each month as the denominator. Rate ratios were used to more other medications in addition to SMC drugs, com- compare rates between surveillance methods, age groups, prising 53 who had taken camphorated tincture of opium and calendar months, estimated using Poisson regression (paregoric, a diarrhea treatment including 0.4 mg/mL with a robust standard error to account for clustering within morphine), 49 who had taken metopimazine (an antie- health posts and with stratification by district. The mean metic), 43 who had taken paracetamol, 24 who had taken number of symptoms reported per patient was compared zinc, 7 who had taken phloroglucinol (a treatment for with Poisson regression. To assess whether particular abdominal pain), 2 who had taken amoxicillin, 2 who had 194 J. A. Ndiaye et al. taken mequitazine (an antihistamine), 2 who had taken Rate per 1000 subjects per month metronidazole (an antibiotic used to treat protozoal infec- Naonal system tion), 1 who had taken chloroquine, and 1 who had taken Enhanced spontaneous reporng Acve surveillance traditional medicine. None of the events reported were considered serious. 3.1 Enhanced Spontaneous Reporting Overall, 1145 events were reported over 3 months, a rate of 30.6 [95% confidence interval (CI) 28.8–32.4] per 1000 children treated per month, compared with a rate of 1.65 (95% CI 1.27–2.15) per 1000 per month in health posts using the national system (Table 2). A total of 927 events (81% of the total) occurred within 10 days of the start of Fig. 1 Average incidence of each symptom, as reported through each the SMC cycle. The incidence of AEs decreased in each method of surveillance successive month, from 30.1/1000 in September, to 25.2 in October and 10.1 in November. The incidence was slightly SMC drugs, 27% of AEs were considered probably related lower in infants than in older children (Table 4). The most to SMC, 36% were possibly related, 1.2% were unlikely to commonly reported symptoms were fever, vomiting, and be related, and causality could not be assessed for the abdominal pain (Fig. 1). Among the older children remaining 36%. (5–10 years of age), the most common symptoms were, in In health posts in the same district that used the national descending order, fever, vomiting, abdominal pain, head- system of reporting, there were 55 events, a rate of 1.65/ ache, and diarrhea (Fig. 2), with a similar pattern being 1000/month. Rates similarly declined with each successive observed in the 12–59 months age group. Among children month. No AE was considered serious. When reports were aged 3–11 months, the most common symptoms were assessed for their association with SMC drugs, 11% were fever and diarrhea. Tiredness, cough, loss of appetite, considered probably related to SMC, 81% possibly related, dizziness, and pruritus were less commonly reported, and 1.2% unlikely to be related, and 6.3% could not be other symptoms (Electronic Supplementary Table S4) classified. together accounted for \ 1% of reported symptoms. No Enhanced reporting involving CHWs and using mobile AE was considered serious. The average number of phones increased reporting 18-fold (rate ratio 18.5, 95% CI symptoms reported per event was 1.67, with 50% of 8.65–39.7), and the incidence of episodes of vomiting by patients reporting more than one symptom. The distribution 10-fold (rate ratio 10.2, 95% CI 5.8–18.0). of the number of symptoms per event is shown in elec- tronic supplementary Table S2. Symptoms varied by age 3.2 Active Surveillance group (Fig. 2). Vomiting was reported at a rate of 10.7/ 1000/month (Table 3), and was most commonly associated Overall, 680 events were reported over 3 months, corre- with diarrhea or fever (electronic supplementary Table S3). sponding to a rate of 21.6 (95% CI 20.1–23.3) per 1000 When the reports were assessed for their relationship with children treated per month, compared with a rate of 3.24 Table 2 Incidence of adverse events (reports with one or more symptoms) following SMC distribution, using three methods of surveillance No. of events No. of treatments Incidence rate/1000 (95% CI) Incidence rate ratio (95% CI) Kolda National system 103 31,818 3.2 (2.7–3.9) Reference Active surveillance 680 31,452 21.6 (20.1–23.3) 6.7 (1.3–33.9) Sedhiou National system 55 33,333 1.65 (1.3–2.1) Reference Enhanced spontaneous reporting 1145 37,458 30.6 (28.8–32.4) 18.5 (8.65–39.7) SMC seasonal malaria chemoprevention, CI confidence interval Incidence rate ratios comparing the surveillance methods were estimated using Poisson regression, with the estimated number of SMC treatments as offset, and using robust standard errors to account for clustering within health post Community-Based Safety Monitoring During Seasonal Malaria Chemoprevention Campaigns in Senegal 195 Fig. 2 Incidence of adverse events, by age group, for each surveillance method. a Enhanced spontaneous reporting; b active surveillance Table 3 Incidence of vomiting following SMC distribution, using three methods of surveillance No. of events No. of treatments Incidence rate/1000 (95% CI) Incidence rate ratio (95% CI) Kolda National system 47 31,818 1.5 (1.1–2.0) Reference Active surveillance 335 31,452 10.7 (9.6–11.9) 7.2 (1.8–29.1) Sedhiou National system 34 33,333 1.0 (0.73–1.4) Reference Enhanced spontaneous reporting 389 37,458 10.4 (9.4–11.5) 10.2 (5.8–18.0) SMC seasonal malaria chemoprevention, CI confidence interval (95% CI 2.67–3.93) per 1000 per month in health posts percent of these reports were considered probably related using the national system. The incidence of AEs decreased to SMC, 70% were possibly related, while the remaining in each successive month, from 30.1/1000 in September, to 10% could not be classified. Active surveillance increased 25.2 in October and 10.1 in November. The incidence rate reporting almost sevenfold (rate ratio 6.7, 95% CI increased with increasing age, from 12.1/1000/month in 1.3–33.9) and reporting of episodes of vomiting by sev- infants, to 21.5 in children 12–59 months of age, and 24.0 enfold (rate ratio 7.2, 95% CI 1.8–29.1). in older children (Tables 4 and 5). The average number of symptoms reported per event was 1.2, with only one symptom reported in the majority (79%) of events. Among 4 Discussion the older children (5–10 years of age), the most common symptoms were, in descending order, vomiting, abdominal Although efforts have been made to strengthen national pain, fever, diarrhea, and headache. Vomiting was reported pharmacovigilance capacity during SMC programs [10], it at a rate of 10.7/1000/month. Other symptoms included is recognized that national systems based on spontaneous itching, lethargy, cough, loss of appetite, and dizziness, but reporting tend to underreport events. There might be these were uncommon. No AE was considered serious. multiple reasons for this: (1) events may not be recognized When reports were assessed for their association with SMC by the health worker as potentially related to administration drugs, 8.6% were considered probably related to SMC, of the medicine; (2) health staff may be too busy to report 72% possibly related, 0.24% were unlikely, and 19% could the event, or may not know how to report; or (3) the patient not be classified. may seek care outside the formal health system. Mild and In health posts in the same district which used the moderate events may be particularly underrepresented as national system of reporting, 103 events (3.24/1000/month) events may not be considered severe enough by a patient to were reported. No AE was considered serious. Twenty warrant consulting a health worker, or not considered by 196 J. A. Ndiaye et al. Table 4 Incidence of adverse No. of No. of treatments Incidence/1000 Incidence rate ratio (95% CI) events (reports with one or more events symptoms) following SMC distribution, using three Kolda methods of surveillance, by age Active surveillance group and month of Age group (months) administration 3–11 39 3212 12.1 0.56 (0.44–0.70) 12–59 324 15,052 21.5 Reference 60–120 317 13,188 24.0 1.12 (1.10–1.15) Month September 306 10,182 30.1 Reference October 265 10,506 25.2 0.84 (0.35–2.03) November 109 10,764 10.1 0.33 (0.09–1.26) Kolda National system Age group (months) 3–11 4 2842 1.4 0.45 (0.32–0.62) 12–59 48 15,255 3.1 Reference 60–120 51 13,721 3.7 1.16 (0.60–2.24) Month September 77 10,615 7.3 Reference October 19 10,553 1.8 0.25 (0.09–0.68) November 7 10,650 0.7 0.09 (0.03–0.25) Sedhiou Enhanced spontaneous reporting Age group (months) 3–11 114 3491 32.7 1.00 (0.73–1.38) 12–59 557 17,142 32.5 Reference 60–120 474 16,825 28.2 0.85 (0.77–0.94) Month September 570 12,332 46.2 Reference October 406 12,423 32.7 0.71 (0.38–1.30) November 169 12,703 13.3 0.29 (0.24–0.35) Sedhiou National system Age group (months) 3–11 8 2331 3.4 2.57 (1.89–3.48) 12–59 22 16,170 1.4 Reference 60–120 25 14,832 1.7 1.21 (0.46–3.19) Month September 31 10,669 2.9 Reference October 11 11,112 1.0 0.34 (0.24–0.48) November 13 11,552 1.1 0.38 (0.24–0.62) SMC seasonal malaria chemoprevention, CI confidence interval health workers to be important enough to report, but these considerably delayed. More intensive monitoring is there- could become a concern (jeopardizing acceptability or fore required, especially in the early phases of a new public adherence to treatment) if common. A further problem is health program, to establish its safety [6]. that reports are slow to reach national pharmacovigilance In October 2014, a workshop was organized by the and malaria program coordinators, therefore investigation Special Programme for Research and Training in Tropical (and the putting in place of any mitigating actions) may be Diseases (TDR) and the London School of Hygiene & Community-Based Safety Monitoring During Seasonal Malaria Chemoprevention Campaigns in Senegal 197 Table 5 Incidence of vomiting No. of events No. of treatments Rate/1000 (95% CI) Rate ratio (95% CI) following SMC distribution, using three methods of Kolda surveillance, by age group and Active surveillance month of administration Age group (months) 3–11 15 3212 4.67 (2.82–7.75) 0.47 (0.23–0.95) 12–59 147 15,052 9.77 (8.31–11.48) Reference 60–120 173 13,188 13.12 (11.30–15.23) 1.36 (1.10–1.67) Month September 171 10,182 16.79 (14.46–19.51) Reference October 117 10,506 11.14 (9.29–13.35) 0.66 (0.32–1.37) November 47 10,764 4.37 (3.28–5.81) 0.26 (0.07–0.99) Kolda National system Age group (months) 3–11 1 2842 0.35 (0.05–2.5) 0.28 (0.02,3.31) 12–59 19 15,255 1.2 (0.79–1.95) Reference 60–120 27 13,721 2.0 (1.3–2.9) 1.54 (0.51–4.66) Month September 41 10,615 3.9 (2.8–5.2) Reference October 5 10,553 0.47 (0.20–1.14) 0.12 (0.02–0.68) November 1 10,650 0.09 (0.01–0.67) 0.02 (0.00–0.13) Sedhiou Enhanced spontaneous reporting Age group (months) 3–11 15 3491 4.30 (2.59–7.13) 0.39 (0.21–0.70) 12–59 189 17,142 11.03 (9.56–12.72) Reference 60–120 185 16,825 11.00 (9.52–12.70) 0.98 (0.98–0.98) Month September 207 12,332 16.79 (14.65–19.24) Reference October 123 12,423 9.90 (8.30–11.82) 0.59 (0.42–0.82) November 59 12,703 4.65 (3.60–6.00) 0.28 (0.26–0.29) Sedhiou National system Age group (months) 3–11 4 2331 1.72 (0.64–4.57) 2.82 (1.10–7.28) 12–59 10 16,170 0.62 (0.33–1.15) Reference 60–120 20 14,832 1.35 (0.87–2.09) 2.13 (0.89–5.10) Month September 20 10,669 1.88 (1.21–2.91) Reference October 6 11,112 0.54 (0.24–1.20) 0.29 (0.11–0.77) November 8 11,552 0.69 (0.35–1.39) 0.37 (0.15–0.94) SMC seasonal malaria chemoprevention, CI confidence interval Tropical Medicine (LSHTM) to bring together coordinators workshop highlighted the need for operational research of national malaria control programs and national phar- into innovative practical methods that could simplify and macovigilance centers from countries that had introduced expedite reporting and improve promptness of notification SMC, or planned to in the near future, to discuss with of events to the national malaria program and the phar- pharmacovigilance experts approaches that could be used macovigilance center [11]. This led to planning of the to improve safety monitoring of SMC programs. The present study that evaluated two approaches to community- 198 J. A. Ndiaye et al. based pharmacovigilance involving CHWs; one based on would remain, as, in this population, morbidity is seasonal, passive, spontaneous reporting in which CHWs and health with the rainy season being associated with the transmis- facility staff were trained to report AEs using a mobile sion of infections and with malnutrition. In this study, in phone application, and the other approach involving active the enhanced and active surveillance areas, health workers follow-up by CHWs after each SMC campaign to ask about were encouraged to report all events. When reports were AEs that caregivers had been asked to record on a symptom assessed for evidence of a relationship with SMC drugs, card. Both methods involved sensitization of the commu- over one-third could not be reliably classified. Therefore, nity to the importance of reporting suspected AEs, and a the overall rates are likely to overestimate the true rates of central supervision team to process, share, and analyze AEs to SMC drugs, but the rates (or rather the upper reports and to provide feedback to reporters. confidence limits on the rates) can be interpreted as rep- The incidence rate of reported events was higher using resenting an upper bound for the true rate of drug-related community-based spontaneous reporting—31/1000 chil- symptoms. Furthermore, there may have been reporting dren/month, compared with 22/1000 children/month using bias, both by caregivers who may have been less likely to active surveillance. These rates compare with rates of report symptoms not listed on the symptom card, and by 1.7–3.2 through the national system in the same districts. CHWs, whose training emphasized the known side effects The greater incidence in the enhanced spontaneous of SMC drugs. A further limitation of the study (in com- reporting arm of the study reflects the longer period of time mon with routine pharmacovigilance) is that although steps over which events were observed. In the active surveillance were taken to improve laboratory capacity for routine arm of the study, participants were asked about events investigation, it was not possible to routinely perform occurring up to 10 days prior to the visit, whereas any hematological analysis or liver function tests, and therefore event during the month could be reported in the enhanced it is possible that cases of clinically silent neutropenia spontaneous reporting arm. (which can be associated with AQ) and hepatitis (which When specific symptoms were considered, the incidence can be associated with both AQ and SP) may have gone rate of vomiting, which, it is known, can be caused by AQ, undetected. Severe cutaneous reactions, known to be was seen to be similar in the active and enhanced sponta- associated with SP, and extrapyramidal syndrome, known neous reporting arms of the study in infants and young to be caused by AQ, are severe and unmistakable, and the children, and slightly more common among older children fact that no cases of these syndromes were seen suggests it in the active arm of the study, while fever and diarrhea is unlikely that any such events occurred during the study were more commonly reported in the spontaneous report- period in the populations under enhanced or active ing arm of the study. It was hypothesized before the study surveillance. The most commonly reported symptoms were gas- that mild side effects would not be reported spontaneously and active surveillance would be necessary to determine trointestinal disorders (nausea/vomiting, abdominal pain, the true burden. The fact that the incidence of vomiting was diarrhea), fever, headache, tiredness, cough, loss of appe- similar using both methods suggests this is not the case tite, dizziness and pruritus. These symptoms were similar when good access is provided through CHWs based in the to those reported in other studies of antimalarials in chil- village. In the enhanced spontaneous reporting arm of the dren in the Central African Republic [13], Burkina Faso study, 16 CHWs were based in the community, to whom [14], and Senegal [4], and in school children in Uganda patients could report, in addition to the two health post [15]. Fever was more commonly reported in our study nurses. It is likely that this improved access for patients, compared with other studies on SP ? AQ, but this may and the provision of training for these staff to recognize reflect that the definition of fever includes a reported his- AEs and when and how to report, may have contributed to tory of fever, and that the study took place over an the high reporting rate, as has been found in other studies extended period during the rainy season, when febrile ill- [12]. nesses are common. None of the reported AEs was classified as serious. As The incidence of AEs decreased in successive rounds of the number of children who received SMC in the intensi- SMC, a finding that is consistent with the results of others fied surveillance area was at least 23,000, the upper 95% studies [4, 14]. The fact that this effect was observed in confidence limit for the rate of severe AEs per child was 1 active and passive surveillance areas suggests there may in 7800 (i.e. on the basis of this study we can rule out rates have been gradual tolerization to the effects of AQ; greater than this). mothers could also become less likely to report mild A limitation of this study is the lack of suitable controls symptoms over time. to establish the rate of symptoms in children who did not For reports submitted by mobile phone, the average delay receive SMC. Data could have been collected prior to the from case presentation to the notification reaching the cen- start of the SMC campaigns, although some confounding tral office was 24 h. In a study using SMS to monitor AEs Community-Based Safety Monitoring During Seasonal Malaria Chemoprevention Campaigns in Senegal 199 Funding This study was funded by the Special Programme for after immunization in Australia, most patients who received Research and Training in Tropical Diseases. Jean-Louis A. Ndiaye is an SMS query about adverse vaccine reactions sent a supported by a Wellcome Trust Intermediate Fellowship in Public response within 2 h of receiving the query [16]. A study in Health and Tropical Medicine which also funded the open access fee Ghana, where the use of mobile phones to contact patients for this article. was compared with home visits to identify AEs related to Conflict of interest Jean-Louis A. Ndiaye, Ibrahima Diallo, Yous- artemisinin-based combination therapy for treating uncom- soupha NDiaye, Ekoue Kouevidjin, Ibrahima Aw, Fassiatou Tairou, plicated malaria found a slightly higher reporting rate with Tidiane Ndoye, Christine M. Halleux, Isaac Manga, Mbaye Niang mobile phones, and that it was possible to interview the Dieme, Medoune Ndiop, Babacar Faye, Piero Olliaro, Corinne S. Merle, Oumar Gaye, and Paul Milligan declare that they have no majority of patients within a few days [17]. In Uganda, an conflicts of interest. Piero Olliaro, Corinne S. Merle, and Christine M. SMS-based reporting system for monitoring malaria diag- Halleux are staff members of the WHO. The views expressed in this nosis and treatment improved timeliness in data reporting publication are the views of the authors alone and do not necessarily [18]. A pilot study in rural districts of Kenya using mobile represent the decisions, policies, or views of the WHO. text messaging for malaria surveillance demonstrated that Open Access This article is distributed under the terms of the Creative phone reporting can improve timeliness and reporting of Commons Attribution 3.0 IGO License (http://creativecommons.org/ data [19]. Semi-structured interviews and focus group dis- licenses/by/3.0/igo/), which permits unrestricted use, duplication, cussions conducted in conjunction with this study indicated adaptation, distribution, and reproduction in any medium or format, as that active surveillance was appreciated by community long as you give appropriate credit to the original author(s) and the source. In any reproduction of this article there should not be any members, but CHWs found this time-consuming and con- suggestion that WHO or this article endorse any specific organization sidered enhanced spontaneous reporting to be effective; or products. The use of the WHO logo is not permitted. This notice however, technical difficulties in terms of electricity supply should be preserved along with the article’s original URL. and internet connections were noted. The authors intend to publish results from this aspect of the study separately. References 1. World Health Organization. World Malaria Report 2016. http:// www.who.int/malaria/publications/world-malaria-report-2017/ 5 Conclusions report/en/. Accessed 14 Dec 2017. 2. World Health Organization. WHO Policy Recommendation: Sea- This study has shown that involving CHWs in safety sonal Malaria Chemoprevention (SMC) for Plasmodium falci- reporting, and training nurses and CHWs to report using a parum malaria control in highly seasonal transmission areas of the Sahel sub-region in Africa. March 2012 (in English and French). mobile phone application, can be used to enhance safety 2012. http://www.who.int/malaria/publications/atoz/who_smc_ reporting and improve timeliness of notifications, but the policy_recommendation/en/index.html. Accessed 15 Apr 2016. approach relies on training and supervision of CHWs and 3. Phillips-Howard PA, Bjorkman AB. Ascertainment of risk of health facility staff, effective community sensitization, and serious adverse reactions associated with chemoprophylactic antimalarial drugs. Bull World Health Organ. 1990;68(4):493–504. a central team to process reports and provide feedback. 4. NDiaye JL, Cisse B, Ba EH, Gomis JF, Ndour CT, Molez JF, These strategies could be adopted in sentinel sites when et al. Safety of Seasonal Malaria Chemoprevention (SMC) with new public health programs are introduced and scaled up. sulfadoxine–pyrimethamine plus amodiaquine when delivered to Enhanced safety monitoring using this approach could be children under 10 years of age by district health services in Senegal: results from a stepped wedge cluster randomized trial. established in SMC areas; however, for comparison, lab- PLoS One. 2016;11(12):e0168421. oratory investigation should be strengthened and baseline 5. Pal SN, Duncombe C, Falzon D, Olsson S. WHO strategy for incidence should be collected prior to SMC campaigns. collecting safety data in public health programmes: complementing spontaneous reporting systems. Drug Saf. 2013;36(2):75–81. Acknowledgements The authors are grateful to the district medical 6. World Health Organization. The safety of medicine in public teams of Kolda and Sedhiou; the regional medical teams; the com- health programmes: pharmacovigilance an essential tool. Geneva: munities and health staff in the study areas; the National Malaria World Health Organization; 2016. Control Programme, Senegal; the staff of the Centre Anti-Poison, 7. ANSD. Senegal: Enquete Demographique et de Sante Continue Dakar; the Direction de la Pharmacie et des Medicaments, Dakar; and (EDS-Continue) 2016. Rockville, MD. Agence Nationale de la the Faculty of Medecine, University Cheikh Anta Diop. Statistique ertde la Demographie, Daka, Senegal nd the DHS Program, ICF; 2017. Compliance with Ethical Standards 8. World Health Organization. Seasonal malaria chemoprevention with sulfadoxine-pyrimethamine plus amodiaquine in children: a field guide. Geneva: World Health Organization; 2013. http:// Ethical approval All procedures performed involving human par- www.who.int/malaria/publications/atoz/9789241504737/en/. Acces- ticipants were in accordance with the ethical standards of the Ethics Committee in Senegal and with the 1964 Helsinki declaration and its sed 15 Dec 2017. later amendments or comparable ethical standards. Informed consent 9. World Health Organization. The use of the WHO-UMC system was sought from all participants by explaining the aims and activities for standardised case causality assessment. 2012. http://www. prior to seeking signed consent. who-umc.org/Graphics/26649.pdf. Accessed 27 July 2016. 200 J. A. Ndiaye et al. 10. World Health Organization. Integrating pharmacovigilance in dihydroartemisinin–piperaquine compared with sulfadoxine-pyr- Seasonal Malaria Chemoprevention: the story so far. Pharma- imethamine plus amodiaquine for seasonal malaria chemopre- ceuticals Newsletter. 2017;4:33–34. http://apps.who.int/ vention in Burkina Faso. Antimicrob Agents Chemother. medicinedocs/documents/s23296en/s23296en.pdf. Accessed 15 2015;59(8):4387–96. Dec 2017. 15. Nankabirwa J, Cundill B, Clarke S, Kabatereine N, Rosenthal PJ, 11. World Health Organization Special Programme for Research and Dorsey G, et al. Efficacy, safety, and tolerability of three regi- Training in Tropical Diseases (TDR). Strengthening safety mens for prevention of malaria: a randomized, placebo-controlled surveillance in seasonal malaria chemoprevention campaigns in trial in Ugandan schoolchildren. PLOS One. 2010;5(10):e13438. Africa. 2015. http://www.who.int/tdr/news/2015/safety_surv_ 16. Leeb A, Regan AK, Peters IJ, Leeb C, Leeb G, Effler PV. Using mal_chemoprev_camp/en/. Accessed 15 Dec 2017. automated text messages to monitor adverse events following 12. Kabanywanyi AM, Mulure N, Migoha C, Malila A, Lengeler C, immunisation in general practice. Med J Aust. Schlienger R. Experience of safety monitoring in the context of a 2014;200(7):416–8. prospective observational study of artemether–lumefantrine in 17. Kukula VA, Dodoo AA, Akpakli J, Narh-Bana SA, Clerk C, rural Tanzania: lessons learned for pharmacovigilance reporting. Adjei A, et al. Feasibility and cost of using mobile phones for Malar J. 2010;9:205. capturing drug safety information in peri-urban settlement in 13. Djalle ´ D, Njuimo SP, Manirakiza A, Laganier R, Le Faou A, Ghana: a prospective cohort study of patients with uncomplicated Rogier C. Efficacy and safety of artemether ? lumefantrine, malaria. Malar J. 2015;14(1):411. artesunate ? sulphamethoxypyrazine-pyrimethamine and arte- 18. Asiimwe C, Gelvin D, Lee E, Ben Amor Y, Quinto E, Katureebe sunate ? amodiaquine and sulphadoxine- C, et al. Use of an innovative, affordable, and open-source short pyrimethamine ? amodiaquine in the treatment of uncompli- message service-based tool to monitor malaria in remote areas of cated falciparum malaria in Bangui, Central African Republic: a Uganda. Am J Trop Med Hyg. 2011;85(1):26–33. randomized trial. Malar J. 2014;13:9. 19. Githinji S, Kigen S, Memusi D, Nyandigisi A, Wamari A, Muturi 14. Zongo I, Milligan P, Compaore YD, Some AF, Greenwood B, A. Using mobile phone text messaging for malaria surveillance in Tarning J, et al. Randomized noninferiority trial of rural Kenya. Malar J. 2014;13:107.

Journal

Pharmaceutical MedicineSpringer Journals

Published: Jun 1, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off