The role of the Deki Reader™ in malaria diagnosis, treatment and reporting: findings from an Africare pilot project in Nigeria

The role of the Deki Reader™ in malaria diagnosis, treatment and reporting: findings from an... Background: The Deki Reader is a diagnostic device used with rapid diagnostic tests (RDTs) and linked to an online database for real-time uploads of patient information and results. This is in contrast to visual interpretation of malaria RDTs recorded on the District Health Information System (DHIS). This paper compares records for use of the Deki Reader with DHIS records of visual interpretation of RDTs. Results: A total of 4063 patient encounters/tests were recorded on the Deki Reader database between June 1st and December 31st, 2016. These tests were for 2629 persons who presented with fever and had RDT done. In compari- son, data from DHIS 2.0 for same period recorded 7201 persons presenting with fever. 2421 out of the 2629 persons (92.1%), received RDT using Deki Reader compared to 6535 out of 7201 persons (90.4%) recorded on DHIS (p = 0.04). From DHIS records, malaria positivity rate was 51.6% (3375 out of 6535 persons) compared to Deki Reader records of 23.6% (572 out of 2421 persons). The difference between these two rates was significant (p < 0.001). The odds ratio (95% CI) for the association between use of Deki Reader and having a positive malaria result was 0.29 (0.26–0.32). DHIS showed that 4008 persons received Artemisinin-based combination therapy (ACT ) while 3989 persons tested positive with RDT or microscopy, compared to 691 out of 705 persons (98.0%) using Deki Reader. Finally, Deki Reader identified 618 processing and manufacturers errors with an error rate of 15.3%. Conclusion: The Deki Reader is likely a useful tool for malaria diagnosis, treatment, and real-time data management. It potentially improves diagnostic quality, reduces wastage in ACT administration and improves data quality. Background with an effective anti-malarial drug by 2020; and, 100% of Nigeria has a high burden of malaria with all year round health facilities report on key malaria indicators routinely transmission and over 97% of its population at risk of by 2020 [1]. infection [1]. The goal of the National Malaria Elimina - The 2010 Nigerian Malaria Indicator Survey (NMIS), tion Programme (NMEP) is “to reduce malaria burden to put the overall malaria parasite prevalence at 42% detected pre-elimination levels and bring malaria-related mortal- using microscopy [2]. The 2015 NMIS key indicators ity to zero” [1]. In line with this goal, three of its seven showed a reduction in the overall parasite prevalence to objectives are to ensure that all persons with suspected 27.4% detected using microscopy and 45.1% detected using the malaria rapid diagnostic test (RDT) [3]. This reduction malaria who seek care are tested with RDT or microscopy is likely attributable to an intensification of malaria con by 2020; all persons with confirmed malaria seen in pri - - vate or public health facilities, receive prompt treatment trol and eradication efforts over the period. Underpinning effective implementation of point-of-care (PoC), testing with RDT, is the programmatic paradigm shift from pre- *Correspondence: omosivie.maduka@uniport.edu.ng sumptive diagnosis and treatment of all fevers as malaria Department of Preventive and Social Medicine, College of Health to a parasite-based diagnosis of malaria with microscopy Sciences, University of Port Harcourt, PMB 5323, Port Harcourt, Nigeria Full list of author information is available at the end of the article © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/ publi cdoma in/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Adah et al. Malar J (2018) 17:221 Page 2 of 10 or RDTs, and treatment with artemisinin-based combina- improve testing, diagnosis and reporting of malaria using tion therapy (ACT) in line with the World Health Organi- RDT in Primary Health Care Centres (PHCs) in two zation’s 2010 and 2015 guidelines [4, 5]. states in the Niger Delta region of Nigeria. This paper The role of RDTs in malaria control efforts has been presents the findings of a comparison between monthly well documented [6–8]. RDTs for malaria began to be aggregated Fionet records and DHIS records for fever available on a large-scale in Nigeria in 2010 [9, 10]. While cases, malaria diagnosis, and malaria treatment. acceptance has risen in recognition of RDT as a reliable and cost-effective test for parasite-based diagnosis of Methods malaria, there have been challenges. These challenges This was a retrospective records-based study to compare include concerns about the ability of RDTs to detect fever cases, malaria diagnosis and malaria treatment at low levels of parasitaemia, and the perceived subjectiv- the study Primary Health Care Facilities (PHCs). ity involved in the interpretation of the results [6, 8, 11]. These challenges have resulted in instances of rejection of RDT results, and treatment for malaria in the presence Study sites and population of a negative RDT result [1, 9, 10, 12]. In the light of this, Africare piloted the use of the Deki Reader and Fionet innovations which improve the accuracy of RDT based technology in two states of the Niger Delta region of testing and diagnosis will be beneficial to malaria control Nigeria; Rivers and Akwa-Ibom. These are states where efforts. Africare had received funding for project implementa- Nigeria’s health system faces significant challenges in tion. Thirty Deki Readers were deployed in 30 PHCs data reporting for decision-making. The District Health across four LGAs: Eket and Ibeno in Akwa Ibom State, Information System (DHIS) version 1.4 was rolled out and Ogu Bolo and Bonny Island in Rivers State. The 30 in 2006 with migration to version 2.0 in 2012 [13, 14]. PHCs selected where those which had the highest turno- DHIS is used for routine reporting of health informa- ver of clients over the preceding 6 months. tion at all levels of healthcare delivery in Nigeria. Malaria The functions of the Deki Reader include: programme indicators are integrated into the DHIS plat- form. Currently, DHIS relies heavily on health workers 1. Automated interpretation of RDTs using image anal- manually entering data on a daily and monthly basis using ysis software; the Health Management Information System (HMIS) 2. Digital data capture, using a touch-screen and a sim- registers at the facility level. Also, it relies on Local Gov- ple user interface software; ernment Monitoring and Evaluation (M&E) officers’ 3. Transmission in real time using local mobile phone collating data from facilities and entering it into the elec- network of processed RDT image, diagnostic event tronic DHIS 2.0 platform at the local government level. data collected, geo-positioning of the device, and The concurrent use of manual and electronic reporting date and time stamp to a central database, which is increases the likelihood of error. There is also the issue accessible via the internet. of delays in availability of data for decision-making using the DHIS platform. This is because data from the index Description of Deki Reader records month is uploaded at the beginning of the new month at One Deki Reader was deployed to each of the 30 selected the earliest, and by the middle of the new month at the PHCs in the two states. Each RDT test done was denoted latest [14–18]. as a patient encounter. For each patient encounter, patient The Deki Reader is an in  vitro diagnostic device used information, test results and images of each test cassette with commercially available rapid diagnostic tests and were recorded in a Deki Reader for real-time upload onto Fionet mobile software (www.fio.com). The Deki Reader the Fionet database. This database was made available provides: step-by-step guidance for performing rapid via password access to malaria programme officers, and diagnostic tests, quality checks for rejecting wrongly pro- select Africare staff. cessed tests, an objective analysis of test results, test-by- test traceability via records uploaded to Fionet, feedback from remote managers using Fionet two-way messaging Description of RDTs used and configurable workflows for standardizing care deliv - ™ All pilot facilities use CareStart Malaria HRP2 (Pf) ery and data capture [19]. brand of RDTs manufactured by AccessBio. These are The Deki Reader has successfully been deployed in RDTs that have high specificity and sensitivity for the Kenya [16], Tanzania [20], and Colombia [21]. In May diagnosis of malaria infection from Plasmodium falci- 2016, Africare, an international NGO deployed 30 Deki parum using whole blood of patients. Results are ready Readers and Fionet mobile software in a pilot project to within 20 min. Adah et al. Malar J (2018) 17:221 Page 3 of 10 Description of comparison data from DHIS above. Almost two-thirds (62.6%) of patient encounters In addition to the use of the Deki Reader to capture were for women, of which 23.7% were pregnant (Table 1). patient information and interpret RDT tests, the staff at each pilot facility continued to conduct visual interpreta- Malaria diagnostic testing tion of RDT results and routinely capture patient infor- Between June and December 2016, 2629 persons who mation on facility registers and the DHIS platform. From presented with fever had their details recorded on Deki this platform, the relevant data was extracted by Africare Readers at the pilot facilities in comparison to 7201 per- programme officers. This was used as comparison data in sons with fever whose details were recorded in the facility this study. registers and the DHIS 2.0. This implies that only 36.5% of persons who presented with fever at the health facili- ties were offered an RDT using Deki Reader assuming Data collection and analysis DHIS 2.0 data captured all patients. Data was sourced from the DHIS 2.0 and Fionet database ® ® Among those who presented with fever and were and collated in an MS Excel spreadsheet. WinPepi sta- offered an RDT using Deki Reader, a total of 2421 out tistical software [22] was used to carry out descriptive of 2629 (92.1%) persons, received diagnostic testing for and inferential statistics to compare data from DHIS and malaria with an RDT. In comparison, 6535 out of 7201 Fionet. Outcome variables for comparison included the persons (90.4%) were recorded on the DHIS as having proportion of fever cases tested for malaria using RDTs, received diagnostic testing for malaria using RDT. The malaria positivity rates, and proportion of persons with overall proportion of cases tested was significantly differ - positive RDT treated for malaria with the recommended ent between Fionet and DHIS 2.0 data (Chi square = 4.25, ACT. The test of significance used was a Chi square to p value = 0.04). However, analysis of the proportions of test the homogeneity of the data, with alpha value set at individuals tested for each month revealed variation in 0.05. Odds ratio was also computed to measure the asso- proportions of persons with fever who received RDT ciation between the use of the Deki Reader and malaria diagnostic testing across months. Data from Fionet had positivity rates. the proportion of cases tested ranging from 88 to 98%. DHIS data had the proportion of cases tested ranging Ethical considerations from 84 to 97% (Table 2). Ethical approval was obtained from the Research Eth- There was a significant amount of variation in the ics Committee of the University of Port Harcourt. Also, number of records between the sources, with twenty- Africare maintains a working agreement with the Rivers five of the thirty facilities having over 20% difference in and Akwa Ibom State Malaria Elimination Programmes the number of patients recorded. The number of records (SMEPs). Data was kept secure, and only authorised per- by the facility is shown in Fig.  1, while the proportion of sonnel within SMEP and Africare were allowed access. patients tested positive by facility is shown in Fig 2. Results Malaria positivity rates A total of 4063 patient encounters were recorded on the Of the 6535 persons with fever tested for malaria between Deki Reader between the June 1st and December 31st, June and December 2016 recorded on the DHIS, 3375 2016. Of these, 2629 persons presented with fever. Major- representing 51.6%, tested positive for malaria. In con- ity of visits (78.9%) were from patients aged 5  years and trast, Deki Reader records show that 572 persons tested Table 1 Baseline characteristics of patient encounters with Deki Reader at pilot PHCs between June and December 2016 Month Under 5 5 and above Male Female Pregnant Nonpregnant Total June 116 556 261 411 74 337 672 July 70 401 149 322 81 241 471 August 96 563 286 373 81 292 659 September 97 444 178 363 100 263 541 October 163 478 246 395 96 299 641 November 206 444 252 398 82 316 650 December 110 319 147 282 88 194 429 Total 858 3205 1519 2544 602 1942 4063 This refers to entries into the Deki Reader not the number of persons seen in the health facility Adah et al. Malar J (2018) 17:221 Page 4 of 10 Table 2 Comparing Deki Reader and DHIS data for fever cases tested with an RDT in Rivers and Akwa Ibom Pilot facilities Month Data from Fionet Data from DHIS 2.0 Chi square (p-value) Number Fever cases Proportion of fever Number Fever cases Proportion of fever of persons tested cases tested of persons tested cases tested with fever with RDT with RDT (%) with fever with RDT with RDT (%) June 500 466 93.2 1056 904 85.6 8.59 (<.0.001) July 247 243 98.4 948 825 87.0 26.60 (<.0.001) August 293 268 91.5 771 727 94.3 2.80 (0.095) September 368 330 89.7 668 611 91.5 0.92 (0.34) October 437 423 96.8 1452 1225 84.4 46.63 (<.0.001) November 452 397 87.8 1245 1195 95.9 37.97 (<.0.001) December 332 294 88.6 1061 1027 96.8 35.04 (<.0.001) Total 2629 2421 92.1 7201 6535 90.4 4.25 (0.04) positive out of 2421 with fever, representing a 23.6% as adding too much blood (1.1%) and not analysing RDTs malaria positivity rate. The difference between these in the appropriate amount of time (4.5%). Other reasons proportions was significant (Chi square = 562.63, for these processing errors are described in Fig.  3. Some p-value < 0.001). Furthermore, the odds ratio (95% CI) select Deki Reader images captured from four pilot facili- for the association between use of Deki Reader and hav- ties showing ‘control line too low’ can be seen in Fig. 4. ing a malaria positive result was 0.29 (0.26–0.32). When analysing the data by facility, the DHIS 2 data shows sev- Discussion enteen, out of thirty, facilities had positivity rates over The results of this study show the potential of the Deki 33% (highlighted in red). On the other hand, Fionet had Reader to improve facility-based malaria indicators an overall positivity rate of 23% with only five facilities related to program objectives. Study findings revealed having positivity rates over 33% (highlighted in red). This significant disparities between the number of clients indicates that the higher positivity rates observed with recorded via the facility registers for monthly upload- DHIS 2.0 was driven by a few facilities reporting high ing onto the DHIS 2.0 platform and the real-time data of positivity rates (Table 3). patient encounters generated from entries into the Deki Reader. Only about a third of records entered into the Treatment rates DHIS had RDTs done using Deki Reader. Malaria posi- According to facility-based DHIS records, a total of 4008 tivity rates using the Deki Reader was about half of that persons received ACT between June and December 2016 obtained from the DHIS database with significantly lower in the facilities. This translates to 101.0% as proportion of odds of having a malaria positive result if the RDT was persons testing positive for malaria who received malaria done with a Deki Reader, compared with RDT results treatment. Furthermore, in June, July, September, and recorded on DHIS 2.0. The positivity rate by facility var - November 2016, more persons received ACT than were ied substantially in the DHIS 2.0 data. The number of tested for malaria using either RDT or microscopy. In persons who received ACTs as recorded by DHIS 2.0 was contrast, the results from the Deki Reader database show much higher than in the Deki Reader data. There was that 690 out of 705 (98.0%) persons who tested positive greater variability in the proportion of persons treated for malaria using a Deki Reader received ACT (Table 4). with ACT recorded on DHIS 2.0 data with instances of treatment rates above 100%. Finally, the Deki Reader was Detecting RDT errors useful in identifying various RDT manufacturing and Between June and December 2016, Deki Reader identi- processing errors in over a tenth of RDTs done with the fied a total of 618 processing and manufacturers’ errors highest error rates coming from no control line on the with error rate of 15.3%. The major error identified was RDT. “control line too low,” representing 405 errors (10.6%). These findings have several implications for facility- This error could be due to a manufacturing defect or based malaria diagnosis, treatment and information inadequate amount of buffer being used so that there management systems. The disparity between the number is not enough buffer to reach the control line. Process - of consultations and diagnostic tests recorded on Fionet ing errors were primarily caused by mistakes made by as compared to DHIS may imply that staff are still adjust - the user of the Deki Reader in processing of RDT such ing to the perceived additional workload of using the Adah et al. Malar J (2018) 17:221 Page 5 of 10 Fig. 1 Number of records by facility of the number of patients tested with malaria RDT over the study period Deki Reader or are making selective decisions about for advantages of automated data generation and upload. whom to use the Deki Reader. Additionally, each facility Experimental and qualitative studies may be useful in had only one Deki Reader with possibly multiple RDT reconciling the observed disparity. Supportive supervi- stations. Issues with acceptability of the location of the sion aimed at identifying and resolving bottlenecks to the Deki Reader, unreliable electricity supply, may also be use of the Deki Reader will also be beneficial. an inhibitor to Deki Reader uptake in these facilities. On Study findings showed a disparity between malaria the other hand, the observed disparity may be a pointer positivity rates from Deki Reader and DHIS data. Fur- to issues with DHIS data integrity, especially as DHIS is thermore, using the Deki Reader ensured decreased odds heavily dependent on retrospective data entry and col- of having a positive result for malaria. This points to the lation by facility and LGA staff in comparison with the Deki Reader device as a potentially useful intervention Deki Reader, which uploads records through mobile for improving the accuracy of RDTs. Figures obtained networks to a web-based portal. This underscores the with Deki Reader closely approximate positivity rates Adah et al. Malar J (2018) 17:221 Page 6 of 10 Fig. 2 Proportion of patients tested positive by facility. The red bars signify that the positivity rate is above 33% obtained using the gold standard of microscopy during Study findings relating to the high proportions of per - the 2015 Malaria Indicator Survey [3]. The ability of the sons treated with ACT also provide evidence that staff Deki Reader to reduce the likelihood of errors from visual in the pilot PHCs may be dispensing ACT medicines for interpretation may be responsible for the lower malaria malaria treatment to persons who had not been tested positivity rates obtained using Deki Reader in compari- for malaria or even those with negative test results. This son to that recorded on DHIS 2.0. This may imply greater finding also raises questions relating to the quality of data effectiveness of diagnosis-based treatment of malaria uploaded to the DHIS. The Deki Reader requires that the using Deki Reader. post-diagnosis treatment plan be recorded and uploaded in real time. This has the potential to foster the ‘track, test Adah et al. Malar J (2018) 17:221 Page 7 of 10 Table 3 Comparing the  Deki Reader and  visual interpretation (from DHIS data) for  malaria positivity rates in  Rivers and Akwa-Ibom pilot facilities Month Data from Deki Reader/Fionet Data from DHIS 2.0 Chi square Odds ratio (95% (p-value) CI) Fever cases Positive Proportion Fever cases Positive Proportion tested RDT of positive tests tested RDT of positive tests with RDT results (%) with RDT results (%) June 466 115 24.7 904 474 52.4 96.65 (<.0.001) 0.30 (0.23 to 0.38) July 243 51 21.0 825 458 55.5 89.71 (<.0.001) 0.21 (0.15 to 0.30) August 268 83 31.0 727 355 48.8 25.35 (<.0.001) 0.47 (0.34 to 0.64) September 330 80 24.2 611 249 40.8 25.66 (<.0.001) 0.47 (0.34 to 0.63) October 423 93 22.0 1225 783 63.9 222.03 (<.0.001) 0.16 (0.12 to 0.21) November 397 79 19.9 1195 509 42.6 129.37 (<.0.001) 0.22 (0.16 to 0.29) December 294 71 24.2 1027 547 53.3 77.81 (<.0.001) 0.28 (0.21 to 0.38) Total 2421 572 23.6 6535 3375 51.6 562.63 (<.0.001) 0.29 (0.26 to 0.32) Table 4 Comparing Deki Reader and DHIS data for ACT treatment rates in Rivers and Akwa Ibom Pilot Facilities Month Data from Fionet Data from DHIS2.0 Positive RDT Treated Treated with ACT Positive RDT Positive Treated Treated result with ACT (%) result microscopy with ACT with ACT (%) June 135 131 97.4 474 113 613 104.4 July 75 75 100.0 458 128 600 102.4 August 102 98 96.1 355 99 437 96.3 September 105 104 99.0 249 60 336 108.7 October 117 116 99.1 783 82 843 97.5 November 90 89 98.9 509 97 612 101.0 December 81 77 95.1 547 35 567 97.4 Total 705 691 98.0 3375 614 4008 101.0 Describes frequency count of patients who received ACT based on a positive result from RDT or microscopy and treat’ policy. Ultimately, this could be a useful tool The Deki Reader is a relatively new technology. Only to reduce ACT wastages and possibly slow the onset of three other studies have been published on the use of resistance due to irrational use of ACT. The Deki Reader the Deki Reader for malaria diagnosis and testing. also offers prospects for improving data quality regarding These studies highlighted several positive effects of completeness and accuracy. This is because it provides using the Deki Reader for malaria diagnosis, treatment, real-time remote supervision through protocol guidance and patient information management including high and a two-way messaging system including image cap- specificity and sensitivity rates compared to microscopy ture. In contrast, the current M&E system provides little or PCR [16, 20, 21]. Findings from this research are in to no daily oversight for the activities of health workers. agreement with these publications. Research done in Study findings also reveal the role of Deki Readers in 2013 in Tanzania and 2014 in Colombia, showed good reducing the number of invalid tests arising from man- performance using the Deki Reader and highlight its “usefulness in the health care sector” [20, 21]. Research ufacturers’ errors or poor technique of health work- done in 2015 in Kenya, found the Deki Reader to be a ers. Deki Reader captures and uploads images of each paperless innovation that brings promise for “improve RDT test done. This together with its ability to identify - processing errors will aid on-the job-capacity-building ment in quality control and quality assurance of malaria activities for health workers. Of note is the particularly diagnosis, care and data management” [16]. high level of manufacturing errors relating to the con- This study is limited by the retrospective use of trol line being too low. This points to the need for the aggregated data for comparison. A randomized control programme to review the quality assurance processes trial would have provided stronger evidence. These are for procurement, distribution, and storage of RDTs. areas for future research. More research is also needed to explore health worker perspectives about usefulness, Adah et al. Malar J (2018) 17:221 Page 8 of 10 Fig. 3 mRDT error rates for June to December 2016 in Akwa Ibom and Rivers States Fig. 4 Deki Reader images for the most commonly identified error ‘control line too low’ Adah et al. Malar J (2018) 17:221 Page 9 of 10 Rivers and Akwa Ibom State Malaria Elimination Programmes (SMEPs). Data location and challenges with the use of the Deki Reader. was kept secure, and only authorized personnel within SMEP and Africare This study also did not calculate specificity and sen - were allowed access. sitivity of the Deki Reader since a recent study had Funding established this [16]. All pilot facilities use CareStart Exon Mobil Nigeria funded the deployment of the deki reader. However, the brand of RDTs (Fig.  3). As such, study findings relate research itself was funded by the contributing authors. to Deki Reader interpretation of only these RDT. How- ever other studies in which other brands of RDTs were Publisher’s Note in use have shown similar findings [16, 20, 21]. Finally, Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Deki Reader currently creates an entry system that runs parallel to paper-based data collection on HMIS Received: 29 July 2017 Accepted: 14 May 2018 registers for upload to DHIS. There may, therefore, be greater uptake of the Deki Reader when it is the sole system of data capture. References 1. Federal Republic of Nigeria. National malaria strategic plan 2014–2020. Conclusion Abuja: Federal Ministry of Health; 2014. p. 1–134. The Deki Reader is an innovation for evidence-based 2. Nigerian National Population Commission (NPC), National Malaria Control malaria interventions. This study found significant dif - Program (NMCP), ICF International. Nigerian Malaria Indicator Survey 2010. Abuja, Nigeria; 2012. ferences in all malaria program indicators in comparison 3. National Malaria Elimination Programme (NMEP), Nigerian National Popu- with DHIS records for visual interpretation of RDTs. Its lation Commission. Malaria Indicator Survey 2015: key indicators. Abuja, usefulness applies to malaria diagnosis and treatment Nigeria and Rockville, Maryland USA; 2016. 4. WHO. Guidelines for the treatment of malaria, 2nd Edn. Geneva: as well as real-time management of program-specific World Health Organization; 2010. http://whqli bdoc.who.int/publi catio data. It may be relevant for improving diagnostic quality, ns/2010/97892 41547 925_eng.pdf. Accessed 4 Mar 2017. reducing wastage in ACT administration and improving 5. WHO. Guidelines for the treatment of malaria. 3rd ed. Geneva: World Health Organization; 2015. validity of data for decision-making. 6. Reyburn H, Mbakilwa H, Mwangi R. Rapid diagnostic tests compared with malaria microscopy for guiding outpatient treatment of febrile illness in Authors’ contributions Tanzania: randomised trial. BMJ. 2007;334:403. OM, PA and OO designed the research protocol including the data collection 7. Wongsrichanalai C, Barcus M, Muth S, Sutamihardja A, Wernsdorfer WH. A and analysis plan and wrote the first draft of the manuscript. OD, KS, OJ super - review of malaria diagnostic tools: microscopy and rapid diagnostic test vised the data collection and contributed significantly to the data analysis and (RDT ). Am J Trop Med Hyg. 2007;77(Suppl 6):119–27. wrote the second draft of the manuscript. SO, NZ and PU contributed to the 8. Wilson M. Malaria rapid diagnostic tests. Clin Infect Dis. 2012;54:1637–41. final data analysis and the final draft of the manuscript. All authors read and 9. Ezeoke OP, Ezumah NN, Chandler CC, Mangham-Jefferies LJ, Onwujekwe approved the final manuscript. OE, Wiseman V, et al. Exploring health providers’ and community percep- tions and experiences with malaria tests in South-East Nigeria: a critical Author details 1 2 step towards appropriate treatment. Malar J. 2012;11:368. Africare Nigeria, Lagos, Nigeria. Department of Preventive and Social 10. Uzochukwu BSC, Onwujekwe E, Ezuma NN, Ezeoke OP, Ajuba MO, Medicine, College of Health Sciences, University of Port Harcourt, PMB 5323, 3 4 Sibeudu FT. Improving rational treatment of malaria: perceptions and Port Harcourt, Nigeria. Africare USA, Washington DC, USA. Fio Corpora- influence of RDTs on prescribing behaviour of health workers in south- tion, Toronto, Canada. National Malaria Elimination Project (NMEP), Federal east Nigeria. PLoS ONE. 2011;6:e14627. Ministry of Health, Abuja, Nigeria. 11. Uzochukwu BS, Obikeze EN, Onwujekwe OE, Onoka CA, Griffiths UK, Chandler C, et al. Cost-effectiveness analysis of rapid diagnostic test, Acknowledgements microscopy and syndromic approach in the diagnosis of malaria in Nige- The authors acknowledge the National Malaria Elimination Programme, the ria: implications for scaling-up deployment of ACT. Malar J. 2009;8:265. State Malaria Elimination Programmes in Rivers and Akwa Ibom States and 12. Uzochukwu BS, Chiegboka LO, Enwereuzo C, Nwosu U, Okorafor D, Exon Mobil for their technical support for this research. Onwujekwe OE, et al. Examining appropriate diagnosis and treatment of malaria: availability and use of rapid diagnostic tests and artemisinin- Competing interests based combination therapy in public and private health facilities in O.M and P.U declare no competing interests. P.A, O.O, O.D and S.O are staff south-east Nigeria. BMC Public Health. 2010;10:486. of Africare which received support from Exxon Mobil to pilot the use of the 13. Asangansi I, Shaguy J. Complex dynamics in the socio-technical Deki Reader. K.S, O.J, and N.Z are staff of the Fio Corporation which owns the infrastructure: the case of the nigerian health management information patents and rights to the Deki Reader. system. In: Proc 10th int conf soc implic comput dev ctries OR—Dubai Sch Gov. 2009. Availability of data and materials 14. Omole G. Health management information system for decision-making The datasets generated and analysed during the current study are available in Nigeria: challenges and resolutions. Int J Sci Res. 2015;4:2968–74. from Fio Corporation and Nigeria’s District Health Information System (DHIS) 15. Asuzu M. The necessity for a health systems reform in Nigeria. J Com- repositories. The Authors can make these two data sets available on reason- munity Med Prim Health Care. 2004;16:1–3. able request. 16. Soti DO, Kinoti SN, Omar AH, Logedi J, Mwendwa TK, Hirji Z, et al. Feasibility of an innovative electronic mobile system to assist health Consent for publication workers to collect accurate, complete and timely data in a malaria control This is not applicable. programme in a remote setting in Kenya. Malar J. 2015;14:430–517. 17. Reich MR. Introduction to the HSR, Nigeria Issue. Health Syst Reform. Ethics approval and consent to participate 2016;2:273–6. https ://doi.org/10.1080/23288 604.2016.12475 56. Ethical approval was obtained from the Research Ethics Committee of the Uni- versity of Port Harcourt. Also, Africare maintains a working agreement with the Adah et al. Malar J (2018) 17:221 Page 10 of 10 18. Ohiri K, Ukoha NK, Nwangwu CW, Chima CC, Ogundeji YK, Rone A, et al. 21. Herrera S, Vallejo AF, Quintero JP, Arévalo-Herrera M, Cancino M, Ferro An assessment of data availability, quality, and use in malaria program S. Field evaluation of an automated RDT reader and data management decision making in Nigeria. Health Syst Reform. 2016;2:319–30. https :// device for Plasmodium falciparum/Plasmodium vivax malaria in endemic doi.org/10.1080/23288 604.2016.12348 64. areas of Colombia. Malar J. 2014;13:87. 19. Fio Corporation. For rapid testing|better data. Better care. http://fio.com/ 22. Abramson JH. WINPEPI updated: computer programs for epidemiologists, rapid -testi ng/. Accessed 26 May 2016. and their teaching potential. Epidemiol Perspect Innov. 2011;8:1. 20. Shekalaghe S, Cancino M, Mavere C, Juma O, Mohammed A, Abdulla S, et al. Clinical performance of an automated reader in interpreting malaria rapid diagnostic tests in Tanzania. Malar J. 2013;12:141. Ready to submit your research ? Choose BMC and benefit from: fast, convenient online submission thorough peer review by experienced researchers in your field rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Malaria Journal Springer Journals

The role of the Deki Reader™ in malaria diagnosis, treatment and reporting: findings from an Africare pilot project in Nigeria

Free
10 pages
Loading next page...
 
/lp/springer_journal/the-role-of-the-deki-reader-in-malaria-diagnosis-treatment-and-cWGKB0gGke
Publisher
BioMed Central
Copyright
Copyright © 2018 by The Author(s)
Subject
Biomedicine; Parasitology; Tropical Medicine; Infectious Diseases; Entomology; Microbiology; Public Health
eISSN
1475-2875
D.O.I.
10.1186/s12936-018-2356-8
Publisher site
See Article on Publisher Site

Abstract

Background: The Deki Reader is a diagnostic device used with rapid diagnostic tests (RDTs) and linked to an online database for real-time uploads of patient information and results. This is in contrast to visual interpretation of malaria RDTs recorded on the District Health Information System (DHIS). This paper compares records for use of the Deki Reader with DHIS records of visual interpretation of RDTs. Results: A total of 4063 patient encounters/tests were recorded on the Deki Reader database between June 1st and December 31st, 2016. These tests were for 2629 persons who presented with fever and had RDT done. In compari- son, data from DHIS 2.0 for same period recorded 7201 persons presenting with fever. 2421 out of the 2629 persons (92.1%), received RDT using Deki Reader compared to 6535 out of 7201 persons (90.4%) recorded on DHIS (p = 0.04). From DHIS records, malaria positivity rate was 51.6% (3375 out of 6535 persons) compared to Deki Reader records of 23.6% (572 out of 2421 persons). The difference between these two rates was significant (p < 0.001). The odds ratio (95% CI) for the association between use of Deki Reader and having a positive malaria result was 0.29 (0.26–0.32). DHIS showed that 4008 persons received Artemisinin-based combination therapy (ACT ) while 3989 persons tested positive with RDT or microscopy, compared to 691 out of 705 persons (98.0%) using Deki Reader. Finally, Deki Reader identified 618 processing and manufacturers errors with an error rate of 15.3%. Conclusion: The Deki Reader is likely a useful tool for malaria diagnosis, treatment, and real-time data management. It potentially improves diagnostic quality, reduces wastage in ACT administration and improves data quality. Background with an effective anti-malarial drug by 2020; and, 100% of Nigeria has a high burden of malaria with all year round health facilities report on key malaria indicators routinely transmission and over 97% of its population at risk of by 2020 [1]. infection [1]. The goal of the National Malaria Elimina - The 2010 Nigerian Malaria Indicator Survey (NMIS), tion Programme (NMEP) is “to reduce malaria burden to put the overall malaria parasite prevalence at 42% detected pre-elimination levels and bring malaria-related mortal- using microscopy [2]. The 2015 NMIS key indicators ity to zero” [1]. In line with this goal, three of its seven showed a reduction in the overall parasite prevalence to objectives are to ensure that all persons with suspected 27.4% detected using microscopy and 45.1% detected using the malaria rapid diagnostic test (RDT) [3]. This reduction malaria who seek care are tested with RDT or microscopy is likely attributable to an intensification of malaria con by 2020; all persons with confirmed malaria seen in pri - - vate or public health facilities, receive prompt treatment trol and eradication efforts over the period. Underpinning effective implementation of point-of-care (PoC), testing with RDT, is the programmatic paradigm shift from pre- *Correspondence: omosivie.maduka@uniport.edu.ng sumptive diagnosis and treatment of all fevers as malaria Department of Preventive and Social Medicine, College of Health to a parasite-based diagnosis of malaria with microscopy Sciences, University of Port Harcourt, PMB 5323, Port Harcourt, Nigeria Full list of author information is available at the end of the article © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/ publi cdoma in/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Adah et al. Malar J (2018) 17:221 Page 2 of 10 or RDTs, and treatment with artemisinin-based combina- improve testing, diagnosis and reporting of malaria using tion therapy (ACT) in line with the World Health Organi- RDT in Primary Health Care Centres (PHCs) in two zation’s 2010 and 2015 guidelines [4, 5]. states in the Niger Delta region of Nigeria. This paper The role of RDTs in malaria control efforts has been presents the findings of a comparison between monthly well documented [6–8]. RDTs for malaria began to be aggregated Fionet records and DHIS records for fever available on a large-scale in Nigeria in 2010 [9, 10]. While cases, malaria diagnosis, and malaria treatment. acceptance has risen in recognition of RDT as a reliable and cost-effective test for parasite-based diagnosis of Methods malaria, there have been challenges. These challenges This was a retrospective records-based study to compare include concerns about the ability of RDTs to detect fever cases, malaria diagnosis and malaria treatment at low levels of parasitaemia, and the perceived subjectiv- the study Primary Health Care Facilities (PHCs). ity involved in the interpretation of the results [6, 8, 11]. These challenges have resulted in instances of rejection of RDT results, and treatment for malaria in the presence Study sites and population of a negative RDT result [1, 9, 10, 12]. In the light of this, Africare piloted the use of the Deki Reader and Fionet innovations which improve the accuracy of RDT based technology in two states of the Niger Delta region of testing and diagnosis will be beneficial to malaria control Nigeria; Rivers and Akwa-Ibom. These are states where efforts. Africare had received funding for project implementa- Nigeria’s health system faces significant challenges in tion. Thirty Deki Readers were deployed in 30 PHCs data reporting for decision-making. The District Health across four LGAs: Eket and Ibeno in Akwa Ibom State, Information System (DHIS) version 1.4 was rolled out and Ogu Bolo and Bonny Island in Rivers State. The 30 in 2006 with migration to version 2.0 in 2012 [13, 14]. PHCs selected where those which had the highest turno- DHIS is used for routine reporting of health informa- ver of clients over the preceding 6 months. tion at all levels of healthcare delivery in Nigeria. Malaria The functions of the Deki Reader include: programme indicators are integrated into the DHIS plat- form. Currently, DHIS relies heavily on health workers 1. Automated interpretation of RDTs using image anal- manually entering data on a daily and monthly basis using ysis software; the Health Management Information System (HMIS) 2. Digital data capture, using a touch-screen and a sim- registers at the facility level. Also, it relies on Local Gov- ple user interface software; ernment Monitoring and Evaluation (M&E) officers’ 3. Transmission in real time using local mobile phone collating data from facilities and entering it into the elec- network of processed RDT image, diagnostic event tronic DHIS 2.0 platform at the local government level. data collected, geo-positioning of the device, and The concurrent use of manual and electronic reporting date and time stamp to a central database, which is increases the likelihood of error. There is also the issue accessible via the internet. of delays in availability of data for decision-making using the DHIS platform. This is because data from the index Description of Deki Reader records month is uploaded at the beginning of the new month at One Deki Reader was deployed to each of the 30 selected the earliest, and by the middle of the new month at the PHCs in the two states. Each RDT test done was denoted latest [14–18]. as a patient encounter. For each patient encounter, patient The Deki Reader is an in  vitro diagnostic device used information, test results and images of each test cassette with commercially available rapid diagnostic tests and were recorded in a Deki Reader for real-time upload onto Fionet mobile software (www.fio.com). The Deki Reader the Fionet database. This database was made available provides: step-by-step guidance for performing rapid via password access to malaria programme officers, and diagnostic tests, quality checks for rejecting wrongly pro- select Africare staff. cessed tests, an objective analysis of test results, test-by- test traceability via records uploaded to Fionet, feedback from remote managers using Fionet two-way messaging Description of RDTs used and configurable workflows for standardizing care deliv - ™ All pilot facilities use CareStart Malaria HRP2 (Pf) ery and data capture [19]. brand of RDTs manufactured by AccessBio. These are The Deki Reader has successfully been deployed in RDTs that have high specificity and sensitivity for the Kenya [16], Tanzania [20], and Colombia [21]. In May diagnosis of malaria infection from Plasmodium falci- 2016, Africare, an international NGO deployed 30 Deki parum using whole blood of patients. Results are ready Readers and Fionet mobile software in a pilot project to within 20 min. Adah et al. Malar J (2018) 17:221 Page 3 of 10 Description of comparison data from DHIS above. Almost two-thirds (62.6%) of patient encounters In addition to the use of the Deki Reader to capture were for women, of which 23.7% were pregnant (Table 1). patient information and interpret RDT tests, the staff at each pilot facility continued to conduct visual interpreta- Malaria diagnostic testing tion of RDT results and routinely capture patient infor- Between June and December 2016, 2629 persons who mation on facility registers and the DHIS platform. From presented with fever had their details recorded on Deki this platform, the relevant data was extracted by Africare Readers at the pilot facilities in comparison to 7201 per- programme officers. This was used as comparison data in sons with fever whose details were recorded in the facility this study. registers and the DHIS 2.0. This implies that only 36.5% of persons who presented with fever at the health facili- ties were offered an RDT using Deki Reader assuming Data collection and analysis DHIS 2.0 data captured all patients. Data was sourced from the DHIS 2.0 and Fionet database ® ® Among those who presented with fever and were and collated in an MS Excel spreadsheet. WinPepi sta- offered an RDT using Deki Reader, a total of 2421 out tistical software [22] was used to carry out descriptive of 2629 (92.1%) persons, received diagnostic testing for and inferential statistics to compare data from DHIS and malaria with an RDT. In comparison, 6535 out of 7201 Fionet. Outcome variables for comparison included the persons (90.4%) were recorded on the DHIS as having proportion of fever cases tested for malaria using RDTs, received diagnostic testing for malaria using RDT. The malaria positivity rates, and proportion of persons with overall proportion of cases tested was significantly differ - positive RDT treated for malaria with the recommended ent between Fionet and DHIS 2.0 data (Chi square = 4.25, ACT. The test of significance used was a Chi square to p value = 0.04). However, analysis of the proportions of test the homogeneity of the data, with alpha value set at individuals tested for each month revealed variation in 0.05. Odds ratio was also computed to measure the asso- proportions of persons with fever who received RDT ciation between the use of the Deki Reader and malaria diagnostic testing across months. Data from Fionet had positivity rates. the proportion of cases tested ranging from 88 to 98%. DHIS data had the proportion of cases tested ranging Ethical considerations from 84 to 97% (Table 2). Ethical approval was obtained from the Research Eth- There was a significant amount of variation in the ics Committee of the University of Port Harcourt. Also, number of records between the sources, with twenty- Africare maintains a working agreement with the Rivers five of the thirty facilities having over 20% difference in and Akwa Ibom State Malaria Elimination Programmes the number of patients recorded. The number of records (SMEPs). Data was kept secure, and only authorised per- by the facility is shown in Fig.  1, while the proportion of sonnel within SMEP and Africare were allowed access. patients tested positive by facility is shown in Fig 2. Results Malaria positivity rates A total of 4063 patient encounters were recorded on the Of the 6535 persons with fever tested for malaria between Deki Reader between the June 1st and December 31st, June and December 2016 recorded on the DHIS, 3375 2016. Of these, 2629 persons presented with fever. Major- representing 51.6%, tested positive for malaria. In con- ity of visits (78.9%) were from patients aged 5  years and trast, Deki Reader records show that 572 persons tested Table 1 Baseline characteristics of patient encounters with Deki Reader at pilot PHCs between June and December 2016 Month Under 5 5 and above Male Female Pregnant Nonpregnant Total June 116 556 261 411 74 337 672 July 70 401 149 322 81 241 471 August 96 563 286 373 81 292 659 September 97 444 178 363 100 263 541 October 163 478 246 395 96 299 641 November 206 444 252 398 82 316 650 December 110 319 147 282 88 194 429 Total 858 3205 1519 2544 602 1942 4063 This refers to entries into the Deki Reader not the number of persons seen in the health facility Adah et al. Malar J (2018) 17:221 Page 4 of 10 Table 2 Comparing Deki Reader and DHIS data for fever cases tested with an RDT in Rivers and Akwa Ibom Pilot facilities Month Data from Fionet Data from DHIS 2.0 Chi square (p-value) Number Fever cases Proportion of fever Number Fever cases Proportion of fever of persons tested cases tested of persons tested cases tested with fever with RDT with RDT (%) with fever with RDT with RDT (%) June 500 466 93.2 1056 904 85.6 8.59 (<.0.001) July 247 243 98.4 948 825 87.0 26.60 (<.0.001) August 293 268 91.5 771 727 94.3 2.80 (0.095) September 368 330 89.7 668 611 91.5 0.92 (0.34) October 437 423 96.8 1452 1225 84.4 46.63 (<.0.001) November 452 397 87.8 1245 1195 95.9 37.97 (<.0.001) December 332 294 88.6 1061 1027 96.8 35.04 (<.0.001) Total 2629 2421 92.1 7201 6535 90.4 4.25 (0.04) positive out of 2421 with fever, representing a 23.6% as adding too much blood (1.1%) and not analysing RDTs malaria positivity rate. The difference between these in the appropriate amount of time (4.5%). Other reasons proportions was significant (Chi square = 562.63, for these processing errors are described in Fig.  3. Some p-value < 0.001). Furthermore, the odds ratio (95% CI) select Deki Reader images captured from four pilot facili- for the association between use of Deki Reader and hav- ties showing ‘control line too low’ can be seen in Fig. 4. ing a malaria positive result was 0.29 (0.26–0.32). When analysing the data by facility, the DHIS 2 data shows sev- Discussion enteen, out of thirty, facilities had positivity rates over The results of this study show the potential of the Deki 33% (highlighted in red). On the other hand, Fionet had Reader to improve facility-based malaria indicators an overall positivity rate of 23% with only five facilities related to program objectives. Study findings revealed having positivity rates over 33% (highlighted in red). This significant disparities between the number of clients indicates that the higher positivity rates observed with recorded via the facility registers for monthly upload- DHIS 2.0 was driven by a few facilities reporting high ing onto the DHIS 2.0 platform and the real-time data of positivity rates (Table 3). patient encounters generated from entries into the Deki Reader. Only about a third of records entered into the Treatment rates DHIS had RDTs done using Deki Reader. Malaria posi- According to facility-based DHIS records, a total of 4008 tivity rates using the Deki Reader was about half of that persons received ACT between June and December 2016 obtained from the DHIS database with significantly lower in the facilities. This translates to 101.0% as proportion of odds of having a malaria positive result if the RDT was persons testing positive for malaria who received malaria done with a Deki Reader, compared with RDT results treatment. Furthermore, in June, July, September, and recorded on DHIS 2.0. The positivity rate by facility var - November 2016, more persons received ACT than were ied substantially in the DHIS 2.0 data. The number of tested for malaria using either RDT or microscopy. In persons who received ACTs as recorded by DHIS 2.0 was contrast, the results from the Deki Reader database show much higher than in the Deki Reader data. There was that 690 out of 705 (98.0%) persons who tested positive greater variability in the proportion of persons treated for malaria using a Deki Reader received ACT (Table 4). with ACT recorded on DHIS 2.0 data with instances of treatment rates above 100%. Finally, the Deki Reader was Detecting RDT errors useful in identifying various RDT manufacturing and Between June and December 2016, Deki Reader identi- processing errors in over a tenth of RDTs done with the fied a total of 618 processing and manufacturers’ errors highest error rates coming from no control line on the with error rate of 15.3%. The major error identified was RDT. “control line too low,” representing 405 errors (10.6%). These findings have several implications for facility- This error could be due to a manufacturing defect or based malaria diagnosis, treatment and information inadequate amount of buffer being used so that there management systems. The disparity between the number is not enough buffer to reach the control line. Process - of consultations and diagnostic tests recorded on Fionet ing errors were primarily caused by mistakes made by as compared to DHIS may imply that staff are still adjust - the user of the Deki Reader in processing of RDT such ing to the perceived additional workload of using the Adah et al. Malar J (2018) 17:221 Page 5 of 10 Fig. 1 Number of records by facility of the number of patients tested with malaria RDT over the study period Deki Reader or are making selective decisions about for advantages of automated data generation and upload. whom to use the Deki Reader. Additionally, each facility Experimental and qualitative studies may be useful in had only one Deki Reader with possibly multiple RDT reconciling the observed disparity. Supportive supervi- stations. Issues with acceptability of the location of the sion aimed at identifying and resolving bottlenecks to the Deki Reader, unreliable electricity supply, may also be use of the Deki Reader will also be beneficial. an inhibitor to Deki Reader uptake in these facilities. On Study findings showed a disparity between malaria the other hand, the observed disparity may be a pointer positivity rates from Deki Reader and DHIS data. Fur- to issues with DHIS data integrity, especially as DHIS is thermore, using the Deki Reader ensured decreased odds heavily dependent on retrospective data entry and col- of having a positive result for malaria. This points to the lation by facility and LGA staff in comparison with the Deki Reader device as a potentially useful intervention Deki Reader, which uploads records through mobile for improving the accuracy of RDTs. Figures obtained networks to a web-based portal. This underscores the with Deki Reader closely approximate positivity rates Adah et al. Malar J (2018) 17:221 Page 6 of 10 Fig. 2 Proportion of patients tested positive by facility. The red bars signify that the positivity rate is above 33% obtained using the gold standard of microscopy during Study findings relating to the high proportions of per - the 2015 Malaria Indicator Survey [3]. The ability of the sons treated with ACT also provide evidence that staff Deki Reader to reduce the likelihood of errors from visual in the pilot PHCs may be dispensing ACT medicines for interpretation may be responsible for the lower malaria malaria treatment to persons who had not been tested positivity rates obtained using Deki Reader in compari- for malaria or even those with negative test results. This son to that recorded on DHIS 2.0. This may imply greater finding also raises questions relating to the quality of data effectiveness of diagnosis-based treatment of malaria uploaded to the DHIS. The Deki Reader requires that the using Deki Reader. post-diagnosis treatment plan be recorded and uploaded in real time. This has the potential to foster the ‘track, test Adah et al. Malar J (2018) 17:221 Page 7 of 10 Table 3 Comparing the  Deki Reader and  visual interpretation (from DHIS data) for  malaria positivity rates in  Rivers and Akwa-Ibom pilot facilities Month Data from Deki Reader/Fionet Data from DHIS 2.0 Chi square Odds ratio (95% (p-value) CI) Fever cases Positive Proportion Fever cases Positive Proportion tested RDT of positive tests tested RDT of positive tests with RDT results (%) with RDT results (%) June 466 115 24.7 904 474 52.4 96.65 (<.0.001) 0.30 (0.23 to 0.38) July 243 51 21.0 825 458 55.5 89.71 (<.0.001) 0.21 (0.15 to 0.30) August 268 83 31.0 727 355 48.8 25.35 (<.0.001) 0.47 (0.34 to 0.64) September 330 80 24.2 611 249 40.8 25.66 (<.0.001) 0.47 (0.34 to 0.63) October 423 93 22.0 1225 783 63.9 222.03 (<.0.001) 0.16 (0.12 to 0.21) November 397 79 19.9 1195 509 42.6 129.37 (<.0.001) 0.22 (0.16 to 0.29) December 294 71 24.2 1027 547 53.3 77.81 (<.0.001) 0.28 (0.21 to 0.38) Total 2421 572 23.6 6535 3375 51.6 562.63 (<.0.001) 0.29 (0.26 to 0.32) Table 4 Comparing Deki Reader and DHIS data for ACT treatment rates in Rivers and Akwa Ibom Pilot Facilities Month Data from Fionet Data from DHIS2.0 Positive RDT Treated Treated with ACT Positive RDT Positive Treated Treated result with ACT (%) result microscopy with ACT with ACT (%) June 135 131 97.4 474 113 613 104.4 July 75 75 100.0 458 128 600 102.4 August 102 98 96.1 355 99 437 96.3 September 105 104 99.0 249 60 336 108.7 October 117 116 99.1 783 82 843 97.5 November 90 89 98.9 509 97 612 101.0 December 81 77 95.1 547 35 567 97.4 Total 705 691 98.0 3375 614 4008 101.0 Describes frequency count of patients who received ACT based on a positive result from RDT or microscopy and treat’ policy. Ultimately, this could be a useful tool The Deki Reader is a relatively new technology. Only to reduce ACT wastages and possibly slow the onset of three other studies have been published on the use of resistance due to irrational use of ACT. The Deki Reader the Deki Reader for malaria diagnosis and testing. also offers prospects for improving data quality regarding These studies highlighted several positive effects of completeness and accuracy. This is because it provides using the Deki Reader for malaria diagnosis, treatment, real-time remote supervision through protocol guidance and patient information management including high and a two-way messaging system including image cap- specificity and sensitivity rates compared to microscopy ture. In contrast, the current M&E system provides little or PCR [16, 20, 21]. Findings from this research are in to no daily oversight for the activities of health workers. agreement with these publications. Research done in Study findings also reveal the role of Deki Readers in 2013 in Tanzania and 2014 in Colombia, showed good reducing the number of invalid tests arising from man- performance using the Deki Reader and highlight its “usefulness in the health care sector” [20, 21]. Research ufacturers’ errors or poor technique of health work- done in 2015 in Kenya, found the Deki Reader to be a ers. Deki Reader captures and uploads images of each paperless innovation that brings promise for “improve RDT test done. This together with its ability to identify - processing errors will aid on-the job-capacity-building ment in quality control and quality assurance of malaria activities for health workers. Of note is the particularly diagnosis, care and data management” [16]. high level of manufacturing errors relating to the con- This study is limited by the retrospective use of trol line being too low. This points to the need for the aggregated data for comparison. A randomized control programme to review the quality assurance processes trial would have provided stronger evidence. These are for procurement, distribution, and storage of RDTs. areas for future research. More research is also needed to explore health worker perspectives about usefulness, Adah et al. Malar J (2018) 17:221 Page 8 of 10 Fig. 3 mRDT error rates for June to December 2016 in Akwa Ibom and Rivers States Fig. 4 Deki Reader images for the most commonly identified error ‘control line too low’ Adah et al. Malar J (2018) 17:221 Page 9 of 10 Rivers and Akwa Ibom State Malaria Elimination Programmes (SMEPs). Data location and challenges with the use of the Deki Reader. was kept secure, and only authorized personnel within SMEP and Africare This study also did not calculate specificity and sen - were allowed access. sitivity of the Deki Reader since a recent study had Funding established this [16]. All pilot facilities use CareStart Exon Mobil Nigeria funded the deployment of the deki reader. However, the brand of RDTs (Fig.  3). As such, study findings relate research itself was funded by the contributing authors. to Deki Reader interpretation of only these RDT. How- ever other studies in which other brands of RDTs were Publisher’s Note in use have shown similar findings [16, 20, 21]. Finally, Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Deki Reader currently creates an entry system that runs parallel to paper-based data collection on HMIS Received: 29 July 2017 Accepted: 14 May 2018 registers for upload to DHIS. There may, therefore, be greater uptake of the Deki Reader when it is the sole system of data capture. References 1. Federal Republic of Nigeria. National malaria strategic plan 2014–2020. Conclusion Abuja: Federal Ministry of Health; 2014. p. 1–134. The Deki Reader is an innovation for evidence-based 2. Nigerian National Population Commission (NPC), National Malaria Control malaria interventions. This study found significant dif - Program (NMCP), ICF International. Nigerian Malaria Indicator Survey 2010. Abuja, Nigeria; 2012. ferences in all malaria program indicators in comparison 3. National Malaria Elimination Programme (NMEP), Nigerian National Popu- with DHIS records for visual interpretation of RDTs. Its lation Commission. Malaria Indicator Survey 2015: key indicators. Abuja, usefulness applies to malaria diagnosis and treatment Nigeria and Rockville, Maryland USA; 2016. 4. WHO. Guidelines for the treatment of malaria, 2nd Edn. Geneva: as well as real-time management of program-specific World Health Organization; 2010. http://whqli bdoc.who.int/publi catio data. It may be relevant for improving diagnostic quality, ns/2010/97892 41547 925_eng.pdf. Accessed 4 Mar 2017. reducing wastage in ACT administration and improving 5. WHO. Guidelines for the treatment of malaria. 3rd ed. Geneva: World Health Organization; 2015. validity of data for decision-making. 6. Reyburn H, Mbakilwa H, Mwangi R. Rapid diagnostic tests compared with malaria microscopy for guiding outpatient treatment of febrile illness in Authors’ contributions Tanzania: randomised trial. BMJ. 2007;334:403. OM, PA and OO designed the research protocol including the data collection 7. Wongsrichanalai C, Barcus M, Muth S, Sutamihardja A, Wernsdorfer WH. A and analysis plan and wrote the first draft of the manuscript. OD, KS, OJ super - review of malaria diagnostic tools: microscopy and rapid diagnostic test vised the data collection and contributed significantly to the data analysis and (RDT ). Am J Trop Med Hyg. 2007;77(Suppl 6):119–27. wrote the second draft of the manuscript. SO, NZ and PU contributed to the 8. Wilson M. Malaria rapid diagnostic tests. Clin Infect Dis. 2012;54:1637–41. final data analysis and the final draft of the manuscript. All authors read and 9. Ezeoke OP, Ezumah NN, Chandler CC, Mangham-Jefferies LJ, Onwujekwe approved the final manuscript. OE, Wiseman V, et al. Exploring health providers’ and community percep- tions and experiences with malaria tests in South-East Nigeria: a critical Author details 1 2 step towards appropriate treatment. Malar J. 2012;11:368. Africare Nigeria, Lagos, Nigeria. Department of Preventive and Social 10. Uzochukwu BSC, Onwujekwe E, Ezuma NN, Ezeoke OP, Ajuba MO, Medicine, College of Health Sciences, University of Port Harcourt, PMB 5323, 3 4 Sibeudu FT. Improving rational treatment of malaria: perceptions and Port Harcourt, Nigeria. Africare USA, Washington DC, USA. Fio Corpora- influence of RDTs on prescribing behaviour of health workers in south- tion, Toronto, Canada. National Malaria Elimination Project (NMEP), Federal east Nigeria. PLoS ONE. 2011;6:e14627. Ministry of Health, Abuja, Nigeria. 11. Uzochukwu BS, Obikeze EN, Onwujekwe OE, Onoka CA, Griffiths UK, Chandler C, et al. Cost-effectiveness analysis of rapid diagnostic test, Acknowledgements microscopy and syndromic approach in the diagnosis of malaria in Nige- The authors acknowledge the National Malaria Elimination Programme, the ria: implications for scaling-up deployment of ACT. Malar J. 2009;8:265. State Malaria Elimination Programmes in Rivers and Akwa Ibom States and 12. Uzochukwu BS, Chiegboka LO, Enwereuzo C, Nwosu U, Okorafor D, Exon Mobil for their technical support for this research. Onwujekwe OE, et al. Examining appropriate diagnosis and treatment of malaria: availability and use of rapid diagnostic tests and artemisinin- Competing interests based combination therapy in public and private health facilities in O.M and P.U declare no competing interests. P.A, O.O, O.D and S.O are staff south-east Nigeria. BMC Public Health. 2010;10:486. of Africare which received support from Exxon Mobil to pilot the use of the 13. Asangansi I, Shaguy J. Complex dynamics in the socio-technical Deki Reader. K.S, O.J, and N.Z are staff of the Fio Corporation which owns the infrastructure: the case of the nigerian health management information patents and rights to the Deki Reader. system. In: Proc 10th int conf soc implic comput dev ctries OR—Dubai Sch Gov. 2009. Availability of data and materials 14. Omole G. Health management information system for decision-making The datasets generated and analysed during the current study are available in Nigeria: challenges and resolutions. Int J Sci Res. 2015;4:2968–74. from Fio Corporation and Nigeria’s District Health Information System (DHIS) 15. Asuzu M. The necessity for a health systems reform in Nigeria. J Com- repositories. The Authors can make these two data sets available on reason- munity Med Prim Health Care. 2004;16:1–3. able request. 16. Soti DO, Kinoti SN, Omar AH, Logedi J, Mwendwa TK, Hirji Z, et al. Feasibility of an innovative electronic mobile system to assist health Consent for publication workers to collect accurate, complete and timely data in a malaria control This is not applicable. programme in a remote setting in Kenya. Malar J. 2015;14:430–517. 17. Reich MR. Introduction to the HSR, Nigeria Issue. Health Syst Reform. Ethics approval and consent to participate 2016;2:273–6. https ://doi.org/10.1080/23288 604.2016.12475 56. Ethical approval was obtained from the Research Ethics Committee of the Uni- versity of Port Harcourt. Also, Africare maintains a working agreement with the Adah et al. Malar J (2018) 17:221 Page 10 of 10 18. Ohiri K, Ukoha NK, Nwangwu CW, Chima CC, Ogundeji YK, Rone A, et al. 21. Herrera S, Vallejo AF, Quintero JP, Arévalo-Herrera M, Cancino M, Ferro An assessment of data availability, quality, and use in malaria program S. Field evaluation of an automated RDT reader and data management decision making in Nigeria. Health Syst Reform. 2016;2:319–30. https :// device for Plasmodium falciparum/Plasmodium vivax malaria in endemic doi.org/10.1080/23288 604.2016.12348 64. areas of Colombia. Malar J. 2014;13:87. 19. Fio Corporation. For rapid testing|better data. Better care. http://fio.com/ 22. Abramson JH. WINPEPI updated: computer programs for epidemiologists, rapid -testi ng/. Accessed 26 May 2016. and their teaching potential. Epidemiol Perspect Innov. 2011;8:1. 20. Shekalaghe S, Cancino M, Mavere C, Juma O, Mohammed A, Abdulla S, et al. Clinical performance of an automated reader in interpreting malaria rapid diagnostic tests in Tanzania. Malar J. 2013;12:141. Ready to submit your research ? Choose BMC and benefit from: fast, convenient online submission thorough peer review by experienced researchers in your field rapid publication on acceptance support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions

Journal

Malaria JournalSpringer 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