Against Malaria Foundation - May 2015 Update

Published: July 2015

Updates on completed distributions

Major updates on completed distributions are:

• AMF completed two recent distributions: one in Dedza, Malawi and the other in Kasaï Occidental, Democratic Republic of the Congo (DRC).
• We have not yet seen documentation for the Kasaï Occidental distribution (completed in November 2014), which was AMF’s largest distribution to date and its first large-scale distribution outside of Malawi.
• AMF told us that Concern Universal has sent it long-delayed malaria case-rate data for the four districts in Malawi for which AMF has provided nets.

More details below.

Distributions in Malawi

AMF has completed three large distributions (i.e., distributions on the order of hundreds of thousands of nets) in Malawi over the last four years:1

• Ntcheu district, 2011-2012, ~268,000 long-lasting insecticide-treated nets (LLINs): completed; post-distribution check-ups ongoing (more)
• Balaka district, 2013, ~154,000 LLINs: completed; post-distribution check-ups ongoing (more)
• Dedza district, 2014, ~245,000 LLINs: recently completed; 45,000 LLIN mop up distribution remaining (more)

This section provides an update on these distributions.

Dedza 2014 distribution

AMF’s Dedza 2014 distribution, which delivered ~245,000 nets in September-October 2014, has largely been completed.2 However, AMF told us that the pre-distribution survey underestimated the amount of nets needed to adequately cover all households in this region by approximately 45,000 nets.3 AMF has ordered an additional 45,000 nets, and a “mop up” distribution will be implemented in April or May of 2015.4

AMF told us that it is not unusual for the number of nets initially ordered to differ from the number of nets ultimately needed because the number of nets initially ordered is based on a pre-distribution survey, and sometimes these surveys miss households.5 However, AMF said that the difference of 45,000 nets was atypically large; it said the large size of the gap was due to a longer-than-normal delay between the pre-distribution survey and the actual distribution.6 AMF plans to compare the pre-distribution survey data to the distribution survey data to try to understand how such a large gap occurred; it received the data that will allow it to make this comparison in February, but we have not yet asked when the analysis will be completed.7

We have not yet seen a post-distribution report for this distribution; Concern Universal is waiting until after the mop up distribution to complete this report.8 We also have not yet seen photos and video from this distribution.9 AMF told us that photos and videos would be made available for this distribution, but we do not know when they will be available.10

The 6-month post-distribution check-up (PDCU) for the Dedza distribution was scheduled to begin in late March or early April.11 However, AMF told us the PDCU-6 began in late May.12 It said the delay was caused by additional work that took place in the first quarter of 2015 in response to flooding that occurred in January 2015 and that affected 15 of Malawi’s 28 districts.13

Malaria case rate data (relevant to all 3 distributions)

In the distribution agreements for Ntcheu, Balaka, and Dedza, AMF requested that Concern Universal, AMF’s partner in Malawi, provide monthly malaria case rate data from all health centers in the three districts.14 AMF previously told us to expect the results from a data audit in September 2013 and additional malaria case rate data in early 2014.15 Since then, we have seen neither data audits nor the data itself, and AMF has delayed when it expected to share these materials several times.16

In late February, AMF informed us that it had received updated malaria case rate data from Concern Universal.17 AMF told us that this data includes approximately 2 years of malaria case rates from the Ntcheu district (distribution was in 2011-2012), 12 months of rates from the Balaka district (distribution was in 2013), and a few months of rates from the Dedza district (distribution was in late 2014).18 AMF said the data also includes 2 years of case rate information from the Dowa district, where AMF is planning to carry out a distribution in 2015 (more).19

AMF has not yet had time to check the quality of the data; it expected that it would have reviewed and formatted the data by early April, after which it would have shared the results with GiveWell.20 As of writing this update in early April, we have not yet received the data. We plan to follow up on the status of this data in our next AMF update.

Post-distribution check-ups (relevant to all 3 distributions)

The post-distribution check-ups for AMF’s completed distributions in Malawi seem to be proceeding on schedule. Updates on post-distribution check-up data are:

• Data collection for Ntcheu’s 33-month post-distribution check-up was completed in November 2014.21 As of the writing of this update (in early April 2015), the data from this check-up was still being publicly entered on AMF’s website.22
• Data for Balaka’s 12-month post-distribution check-up (conducted in November 2014) was also being publicly entered on AMF’s website.23
• As mentioned above, the 6-month post-distribution check-up for the Dedza 2014 distribution was scheduled to begin by early April 2015.24

Distribution in the DRC

Kasaï Occidental distribution

In November of 2014, AMF completed its distribution of 676,000 nets in the province of Kasaï Occidental in the DRC – its largest distribution to date.25

As of the writing of this update (in early April 2015), we have not yet seen documentation for this distribution, such as:

• Distribution proposal26
• Complete photos and video of the distribution27
• Detailed post-distribution documentation, such as data on the number of households reached and how many nets were distributed28

AMF told us that it has been waiting to receive the distribution report from its distribution partner and that the delay is in part due to staff travel and stretched resources.29 AMF said it plans to work closely with its distribution partner to ensure that such delays are avoided in the future.30 A distribution report and a separate technology report are being produced.31

AMF originally predicted that documentation would be available in mid-March, then later predicted it would be available by the end of March.32 We have the impression that documentation from this distribution is taking longer to be made public than most documentation from AMF’s large-scale distributions in Malawi.33 AMF has told us that this is partially due to a) there being two reports and b) the additional data gathering required for the technology report.34

This distribution was the first that AMF has funded that used smartphones to collect data.35 AMF is preparing a technology report on the use of smartphones; AMF told us that it expected this report to be available by mid-April.36 AMF told us that the initial results were promising, although there were some issues with the use of smartphones.37 For example, some of the smartphones that were ordered turned out to be nonfunctional and during the distribution there were too few back-up batteries.38 AMF said that such problems are to be expected when introducing new technologies and that it plans to continue to experiment with smartphones in future distributions because of their potential benefits.39

The 6-month post-distribution check-up for the Kasaï Occidental distribution will be in May, and AMF said that data from this check-up should be available in June or July.40

Updates on in-progress distributions Major updates on AMF’s in-progress distributions are: Three of AMF’s distributions that were scheduled to take place in late 2014 or early 2015 have been delayed but are expected to go forward in the next several months. Our impression is that the justifications for these delays are broadly reasonable. AMF committed approximately 113,000 nets to a new emergency distribution in Malawi in response to severe flooding. Our impression is that this was a reasonable use of AMF’s resources. Malawi may ask many of the organizations that distribute nets within its borders to implement some of AMF’s distribution practices even in districts for which AMF does not fund distributions. More details below. Distributions in Malawi Dowa 2015 distribution The Dowa 2015 distribution, originally scheduled for January-February 2015, has been delayed, because some of the nets originally intended for this distribution were donated to an emergency distribution (more details below).41 AMF told us that the pre-distribution survey for the Dowa distribution is almost completed and that it expects the distribution to occur sometime during March-May.42 AMF needs to order approximately 113,000 nets to replace those used for the emergency distribution.43 Emergency distribution In early 2015, Malawi was in a state of emergency due to severe flooding.44 AMF told us that it learned that roughly 370,000 people were left unprotected by nets and that malaria rates were beginning to rise.45 AMF had approximately 400,000 nets in a warehouse in Malawi waiting to be distributed in the Dowa district when the Malawi government approached AMF about donating some of those nets.46 AMF agreed to give approximately 113,000 nets, effectively making a new ~$315,000 commitment.47 The nets were to be distributed by 7 different partner organizations in February and March 2015.48 AMF is funding a 3-month post-distribution check-up for the distribution. Concern Universal, AMF’s historical partner in Malawi, will run the check-up.49 For the check-up, 5% of the households that received nets will be randomly visited (the visits will be unannounced) to gather information on whether or not the nets were distributed as claimed and are being used properly.50 AMF told us that monitoring for emergency distributions is rare and that it hopes that its monitoring will set a precedent for future similar situations.51 Balaka, Ntcheu, and Dedza late-2015 to late-2017 distributions The pre-distribution survey for the 2015 Ntcheu distribution (scheduled for October-November 2015) was expected to take place in March and April.52 The pre-distribution survey for the 2015 Balaka distribution (scheduled for September-October 2015) will take place in April and May.53 There are no updates on the Dedza distribution planned for late 2017, as it is fairly far in the future.54 Data collection for other distributions in Malawi AMF primarily distributes nets in 4 of Malawi’s 28 districts.55 However, AMF told us that for the second and third quarter of 2015, the National Malaria Control Program (NMCP) of Malawi is planning to implement some of AMF’s distribution practices in districts for which AMF does not fund distributions.56 AMF told us that AMF staff and Nelson Coelho, an employee of Concern Universal, have worked closely with the NMCP of Malawi to make this happen.57 Additionally, AMF is considering funding an additional staff member to further support the NMCP as it considers the development of the mechanisms it uses to distribute and follow up on nets.58 Some of the practices that will be implemented include: gathering registration data for all households and checking data quality.59 These practices will be implemented in 18 of Malawi’s districts that together will receive 7 million nets in 2015.60 DRC distributions North Idjwi Island and South Idjwi Island The North Idjwi Island distribution of ~62,000 LLINs, originally planned for November 2014 through April 2015, has been delayed due to issues with the pre-distribution survey.61 The pre-distribution survey is intended to only include seven questions. However, AMF told us that the North Idjwi Island officials in charge of the survey decided to add many more questions, which made the survey so long that approximately one-third of the target population refused to take or complete the survey.62 So, the pre-distribution survey has to be re-done, and AMF does not expect to be able to sign off on the distribution until September 2015.63 This means that the distribution will likely not take place until October or November 2015.64 Because the South Idjwi Island distribution (also ~62,000 LLINs) was not going to begin until the North Idjwi Island distribution was completed, the South Idjwi Island distribution, originally planned for April-May of 2015, is also delayed.65 Nord Ubangi distribution The Nord Ubangi distribution is slightly delayed. It was originally scheduled for April through June of 2015.66 However, AMF informed us that the 730,000 nets for the distribution were expected to arrive in July.67 At the time of writing this update (in early April 2015), we have not yet asked AMF why this distribution was delayed. Updates on potential future distributions Major updates on potential future distributions are: Malawi has asked AMF to help fund its universal coverage campaign in 2018. AMF told us that it will likely provide approximately $6 million to support this distribution. AMF is attempting to finalize two new distribution agreements: one for approximately 2.5 million nets in Country X (AMF cannot publicly share the names of countries it is negotiating with), and one for approximately 5.5 million nets over the next three years in Country Y. We do not have a strong sense of how likely it is that AMF will be able to finalize these distribution agreements. AMF is planning to fund a study on insecticide resistance that will be conducted with its 2015 Nord Ubangi distribution. AMF is also in discussion with the NMCPs in several other countries for distributions in 2015 to 2017. AMF has indicated that it could provide at least low millions of nets in each case. More details below. Potential distributions in Malawi In late 2014, AMF was interested in funding additional distributions in Malawi in 2015.68 Since then, AMF has shifted focus to funding additional distributions in Malawi in 2018 because the NMCP of Malawi has requested AMF’s assistance with mass LLIN distributions in 2018.69 AMF is planning to agree to assist Malawi in 2018 in the four districts where AMF has previously provided funding (Ntcheu, Balaka, Dedza, Dowa); AMF estimates that these distributions will cost approximately $6 million.70 Potential distributions in Country X AMF has been working to fund a distribution in Country X.71 Originally, AMF planned to provide approximately $12 million to fund 4 million nets in Country X.72 A co-funding partner planned to provide another approximately $12 million to cover all non-net costs.73 However, AMF told us that the partner that AMF chose to distribute nets in Country X could not come to an agreement with AMF over some of AMF’s monitoring requirements and pulled out of this plan.74 This caused the co-funding partner to also withdraw its funding, because it needed to commit its funds on a specific funding schedule and could not wait for AMF to find a new distribution partner.75 After these events, the head of Country X’s NMCP was able to acquire some funding from other sources, reducing Country X’s need to between 2.5 and 2.7 million nets.76 AMF has continued to offer to fund enough nets to cover the remaining gap in Country X.77 However, AMF told us that it is not in a position to cover the non-net costs of this distribution.78 AMF is currently talking to other actors who may be able to fund the non-net costs.79 Additionally, AMF told us that it has identified another partner for this distribution and has an agreement in place with it to distribute nets if AMF ultimately funds a distribution in Country X.80 AMF has informed Country X that it would pursue other opportunities in case a distribution in Country X could not be worked out.81 Potential distributions in Country Y Large distribution AMF told us that there is currently a need for funding for roughly 9 million nets in Country Y for 2015-2017.82 AMF is currently discussing the possibility of funding 5.5 million nets of this gap.83 A co-funding partner is in place to cover non-net costs.84 AMF told us that it is primarily negotiating its distribution agreement with Country Y’s NMCP.85 AMF said that if they reach an agreement, an implementing distribution partner that agrees to AMF’s monitoring requirements will take responsibility for the net distribution, follow-up and associated reporting.86 Small distribution There is also a smaller distribution of approximately 60,000 nets in Country Y that AMF is interested in funding.87 However, until AMF knows which districts the larger distribution in Country Y may fund, it cannot agree to fund this smaller distribution.88 AMF is particularly interested in this distribution because its implementing partner plans to use smartphones for data collection, and AMF would like to test the use of smartphones for data collection in other contexts.89 Other potential distributions AMF is in discussions with several other countries regarding contributions of a few million nets.90 We do not yet have more details on these potential distributions. Insecticide resistance studies AMF plans to fund at least one, and possibly two, studies on insecticide resistance.91 One of these studies will be conducted by the Liverpool School of Tropical Medicine with the 2015 Nord Ubangi distribution. We have now seen a preliminary research proposal for this study.92 The study will include: Measurements of species abundance, resistance prevalence and level, and the effect of PBO (an insecticide) on insects’ mortality at baseline and at a 1-year post-distribution follow-up93 Monthly measurements of insect pesticide resistance in households covered by different net types (some households will be covered by standard LLINs while others will be covered by standard LLINs that have also been treated with PBO).94 AMF expected to receive the full research proposal in late March, and expected to approve the research plan by mid-April.95 We have not yet followed up with AMF about whether it received the full proposal. Room for more funding It is difficult to assess AMF’s room for more funding because (1) AMF moves money in large chunks once it reaches a large-scale distribution agreement, so AMF would ideally have a large amount of funding ready to commit at any given time, but (2) we have some remaining uncertainty about AMF’s ability to sign new agreements and conduct high-quality large-scale distributions outside of Malawi (more), so we are cautious about directing a large amount of additional funding to AMF. At this point in time, we continue to recommend AMF as an excellent giving opportunity and believe it has room for more funding, but we would be hesitant to recommend that it receive a large amount of additional funding (i.e., on the order of tens of millions of dollars) until we can evaluate results from Kasaï Occidental, its first large-scale distribution outside of Malawi. AMF’s current funding situation As of March 2015, AMF held about $19.6 million.96 Of this amount, about $5.4 million is committed to distributions through the end of 2017, leaving about $14.2 million uncommitted.97 If AMF were to fund all of the potential distributions and projects that it is currently discussing at a high level of detail that would take place in late 2015, AMF would need about $14.6 million: $7.5 million for 2.5 million nets in Country X (assuming $3 per net) $6.6 million for 2.2 million nets in Country Y (assuming $3 per net) $500,000 (estimate) for a study on insecticide resistance during the Nord Ubangi distribution in April98 In addition, AMF may need to commit approximately $6 million within the next six months to Malawi for the 2018 national distribution that Malawi has asked AMF to help fund.99 AMF also hopes to sign an agreement with Country Y that would only hold it legally responsible for providing funds for Country Y’s 2015 distribution, with an implicit agreement that AMF would provide funding for Country Y’s distributions in 2016 and 2017 as well.100 This would give AMF more time to fundraise for those future distributions. Because AMF has limited funding for future potential distributions, it is pursuing a strategy of engaging deeply in discussions with 1-2 countries for large distributions (currently Country X and Country Y), and engaging less deeply with a number of other countries for smaller distributions (in case the large distributions fall through).101 AMF told us that it does not feel constrained by its current staff capacity and does not have any plans to significantly increase its fundraising efforts.102 Sources Document Source 7-partner agreement for emergency distribution (DRAFT) Unpublished AMF All distributions page Source (archive) AMF blog post, February 18, 2015 Source (archive) AMF blog post, March 17, 2015 Source (archive) AMF page on Balaka 2013 distribution Source (archive) AMF page on Dedza 2014 distribution Source (archive) AMF page on Dedza 2017 distribution Source (archive) AMF page on flood affected districts distribution Source (archive) AMF page on future distributions Source (archive) AMF page on Kasaï Occidental 2014 distribution Source (archive) AMF page on non-net costs Source (archive) AMF page on Nord Ubangi 2015 distribution Source (archive) AMF page on Ntcheu 2012 distribution Source (archive) AMF page on the introduction of smartphone technology Source (archive) Balaka (2013) 12-month post-distribution check-up data Source (archive) Concern Universal Balaka 2013 distribution proposal Source (archive) Concern Universal Dedza 2014 distribution proposal Source (archive) Country Y - Population by region Unpublished Country Y 2015 mass LLIN distribution schedule Unpublished Country Y - LLINs gap Unpublished Email attachment from AMF - Country Y 1, sent to GiveWell on February 6, 2015 Unpublished Email attachment from AMF - Country Y 2, sent to GiveWell on February 6, 2015 Unpublished GiveWell 2014 end-of-the-year review of AMF Source GiveWell estimate of AMF cost per net 2014 Source GiveWell’s non-verbatim summary of a conversation with Rob Mather and Peter Sherratt on February 6, 2015 Source Insecticide resistance research proposal Unpublished Malawi 4 Districts Schedule Source Malawi NMCP invitation for 2018 distribution Unpublished Melanie Renshaw, phone conversation with GiveWell, May 23, 2014 Source Ntcheu (2011-12) 33-month post-distribution check-up data Source (archive) Rob Mather, conversation with GiveWell, February 24, 2015 Unpublished Rob Mather and Peter Sherratt, conversation with GiveWell, November 6, 2014 Unpublished Rob Mather and Peter Sherratt, conversation with GiveWell, February 6, 2015 Unpublished Rob Mather, email to GiveWell, February 24, 2015 Unpublished Rob Mather, email to GiveWell, November 6, 2014 Unpublished Rob Mather, email to GiveWell, February 6, 2015 Unpublished Rob Mather, email to GiveWell, March 23, 2015 Unpublished Rob Mather, email to GiveWell, June 19, 2015 Unpublished 1 AMF All distributions page 2 Under the “When” category, the Dedza District distribution for 245,000 nets is listed for “Sep-Oct 14.” AMF page on Dedza 2014 distribution Rob Mather, conversation with GiveWell, February 24, 2015 3 Rob Mather, conversation with GiveWell, February 24, 2015 AMF All distributions page 4 Rob Mather, conversation with GiveWell, February 24, 2015 A mop up distribution gives nets to everyone who did not receive nets in the initial distribution. 5 Rob Mather, conversation with GiveWell, February 24, 2015 6 Rob Mather, conversation with GiveWell, February 24, 2015 Rob Mather, email to GiveWell, June 19, 2015 7 AMF noted that the documents it received were primarily in French, which makes the comparison take slightly longer. Rob Mather, conversation with GiveWell, February 24, 2015 8 “Distribution Summary Report - to follow” under “Documentation” on AMF page on Dedza 2014 distribution Note that we have seen mid-distribution reports from this distribution, which we discuss in this section of our year-end review. Rob Mather, email to GiveWell, June 19, 2015 9 See AMF page on Dedza 2014 distribution, where photos and video would be available. 10 "[GiveWell:] Does AMF expect to post photos/videos from the Balaka distribution? [AMF:] No, CU did not take photos of this one. They are aware of this slip and there will be photos and video footage from Dedza. Photos and videos are also included in the agreement for the DRC-Kasai distribution." Rob Mather, AMF Founder, email to GiveWell, May 22, 2014.” GiveWell 2014 end-of-the-year review of AMF 11 “Dedza, Malawi Mop up nets (45,000) due to be distributed in April or May PDCU-6 to take place next month” Rob Mather, email to GiveWell, February 24, 2015 The above suggests that the 6-month post-distribution check-up will take place before the 45,000 net mop up distribution. Malawi 4 Districts Schedule 12 Rob Mather, email to GiveWell, June 19, 2015 13 Rob Mather, email to GiveWell, June 19, 2015 14 We discuss case rate data from Ntcheu in our February 2013 update on AMF. "[AMF:] Please confirm you are able to provide monthly malaria case rate data going back at least 12 months for each health centre/clinic in the distribution area and will continue to provide monthly data for a period of four years post‐distribution. This ensures we understand pre‐distribution malaria levels and can monitor them post‐distribution. [Concern Universal:] Malaria case rate information for each of the 16 Health Centres is available from the Balaka District Health Office using the Malawi Government’s national Health Management Information System (HMIS). This information will be sent to AMF separately." Concern Universal Balaka 2013 distribution proposal, Pg 2. Concern Universal Dedza 2014 distribution proposal, Pg 2 contains the same language as the above Balaka distribution proposal. 15 “In August 2013, it told us that it planned to publish an audit of the data quality by the end of September 2013. Later, it told us that it expected to post data in early 2014.” GiveWell 2014 end-of-the-year review of AMF 16 “Previously, AMF expected to collect malaria case rate from the regions in which it funded LLIN distributions. AMF has asked Concern Universal to provide monthly malaria case rate data from all health centers in the three districts in which it has completed distributions: Ntcheu, Balaka and Dedza. AMF has asked for this data for 12 months preceding and 4 years following the distribution.22 AMF has shared data from Ntcheu through through December 2012. We do not believe this data is high quality. In August 2013, it told us that it planned to publish an audit of the data quality by the end of September 2013.23 Later, it told us that it expected to post data in early 2014.24 In November 2014, AMF told us that it expects to share more recent data by the end of January 2015. AMF told us it has been waiting to collect data through to September 2014 to be able to review data for a full year and compare it with the prior year due to changes in the way data is recorded in the health centers.25” GiveWell 2014 end-of-the-year review of AMF 17 Rob Mather, conversation with GiveWell, February 24, 2015 18 Rob Mather, conversation with GiveWell, February 24, 2015 19 Since AMF has not yet completed a distribution in the Dowa district, this is merely contextual data. AMF All distributions page Rob Mather, conversation with GiveWell, February 24, 2015 20 Rob Mather, conversation with GiveWell, February 24, 2015 21 Rob Mather, email to GiveWell, February 24, 2015 Data collection was originally intended to be completed in mid-October. See the “Task List - By Distribution” sheet of Malawi 4 Districts Schedule. 22 Ntcheu (2011-12) 33-month post-distribution check-up data 23 Balaka (2013) 12-month post-distribution check-up data 24 “Dedza, Malawi Mop up nets (45,000) due to be distributed in April or May PDCU-6 to take place next month” Rob Mather, email to GiveWell, February 24, 2015 The data collection for the PDCU is scheduled to begin on April 2nd. Malawi 4 Districts Schedule 25 AMF All distributions page AMF page on Kasaï Occidental 2014 distribution 26 “Distribution Proposal - to follow,” under the “Pre-distribution” section on AMF page on Kasaï Occidental 2014 distribution AMF typically posts distribution proposals on its website for each distribution. 27 "[GiveWell:] Does AMF expect to post photos/videos from the Balaka distribution? [AMF:] No, CU did not take photos of this one. They are aware of this slip and there will be photos and video footage from Dedza. Photos and videos are also included in the agreement for the DRC-Kasai distribution." Rob Mather, AMF Founder, email to GiveWell, May 22, 2014.” GiveWell 2014 end-of-the-year review of AMF As of the writing of this update, AMF has posted three videos of eight for the Kasaï Occidental distribution. AMF page on Kasaï Occidental 2014 distribution 28 AMF posts detailed data on number of households reached and nets distributed in public “Post-Distribution Reports” for each distribution. See, for example, the Post-Distribution Reports for its Ntcheu 2012 distribution (AMF page on Ntcheu 2012 distribution, under the “Distribution” page’s “Documentation” section) and its Balaka 2013 distribution (AMF page on Balaka 2013 distribution, under the “Distribution” page’s “Documentation” section). As of the writing of this update (in early April 2015), the AMF page on Kasaï Occidental 2014 distribution “Distribution” “Documentation” section says ”Immediate Post-Distribution Report - to follow.” 29 Rob Mather, email to GiveWell, June 19, 2015 30 Rob Mather, email to GiveWell, June 19, 2015 31 Rob Mather, email to GiveWell, June 19, 2015 32 Originally, AMF predicted the documentation would be ready mid-March. Rob Mather, conversation with GiveWell, February 24, 2015 However, AMF later predicted it should be ready by the end of March Rob Mather, email to GiveWell, March 23, 2015. 33 See this section for details on what types of documentation we expect to see from the Kasaï Occidental distribution. Since distribution proposals should be created before a distribution, we would expect to see the proposals posted before or during a distribution. We do not have a strong sense of when we should expect photos and videos or post-distribution reports to be publicly available. For the Ntcheu 2012 distribution, a post-distribution report was publicly available roughly a month after the distribution was completed. For the Balaka 2013 distribution, which was completed in December 2013, it appears that the post-distribution report was published by at least March 2013 since there is a “Note” in the report from March 2013. For the Dedza 2014 distribution, AMF provided mid-distribution reports during the distribution, but it has not yet published a post-distribution report (perhaps because it still plans to distribute an additional 45,000 nets in Dedza, as noted above). It has now been roughly five months since the Kasaï Occidental distribution was completed. See AMF page on Ntcheu 2012 distribution, AMF page on Balaka 2013 distribution, and AMF page on Dedza 2014 distribution. 34 Rob Mather, email to GiveWell, June 19, 2015 35 “Our distribution of 676,000 nets in Kasaï Occidental in partnership with IMA World Health (IMA) is our first one using smartphone technology for data collection.” AMF page on the introduction of smartphone technology 36 Rob Mather, conversation with GiveWell, February 24, 2015 We plan to follow up on this report in our next AMF update. 37 Rob Mather, email to GiveWell, June 19, 2015 38 Rob Mather, conversation with GiveWell, February 24, 2015 39 Rob Mather, conversation with GiveWell, February 24, 2015 ”This distribution is the first that both parties have undertaken using smartphones to collect household-level data. We expect teething problems, as with any new introduction of technology, especially in a country such as DRC. However, we are undeterred by this, as the benefits of improving processes in this region are so significant.” AMF blog post, February 18, 2015 40 Rob Mather, conversation with GiveWell, February 24, 2015 41 AMF All distributions page Rob Mather, conversation with GiveWell, February 24, 2015 42 Rob Mather, conversation with GiveWell, February 24, 2015 Rob Mather, email to GiveWell, February 24, 2015 43 Rob Mather, conversation with GiveWell, February 24, 2015 44 Rob Mather, conversation with GiveWell, February 24, 2015 45 Rob Mather, conversation with GiveWell, February 24, 2015 46 Rob Mather, conversation with GiveWell, February 24, 2015 47 “Total funding required for nets: US$315,348” AMF page on flood affected districts distribution 48 Rob Mather, conversation with GiveWell, February 24, 2015 7-partner agreement for emergency distribution (DRAFT) 49 Rob Mather, conversation with GiveWell, February 24, 2015 7-partner agreement for emergency distribution (DRAFT) 50 Rob Mather, conversation with GiveWell, February 24, 2015 7-partner agreement for emergency distribution (DRAFT) 51 Rob Mather, conversation with GiveWell, February 24, 2015 52 “Ntcheu, Malawi PDCU-33 months completed Nov14 PDRS to take place in March/April” Rob Mather, email to GiveWell, February 24, 2015 53 “Balaka, Malawi PDRS to take place in April/May15” Rob Mather, email to GiveWell, February 24, 2015 54 AMF page on Dedza 2017 distribution 55 ”Malawi is composed of three regions (the Northern, Central and Southern regions),[5] which are divided into 28 districts,[32] and further into approximately 250 traditional authorities and 110 administrative wards.[5]” Malawi Wikipedia page The districts in which AMF has completed or planned distributions are: Ntcheu, Balaka, Dedza, and Dowa. AMF All distributions page 56 Rob Mather, conversation with GiveWell, February 24, 2015 ”In Malawi, the NMCP has adopted some of AMF methodologies to do with data collection and verification in its upcoming 7 million net, 18- district distribution in 2015. Although AMF won’t be managing this process, it is interested to observe how multiple districts will carry out distribution according to this particular data-driven approach. AMF believes that it will be allowed access to that data, either directly or via its in-country partners.” GiveWell’s non-verbatim summary of a conversation with Rob Mather and Peter Sherratt on February 6, 2015 57 Rob Mather, conversation with GiveWell, February 24, 2015 58 Rob Mather, email to GiveWell, June 19, 2015 59 Rob Mather, email to GiveWell, February 24, 2015 60 Rob Mather, email to GiveWell, June 19, 2015 61 Rob Mather, conversation with GiveWell, February 24, 2015 See the North Idjwi Island, South Kivu distribution: AMF All distributions page 62 Rob Mather, conversation with GiveWell, February 24, 2015 63 AMF will not approve the distribution of nets until it has seen the pre-distribution survey. Rob Mather, conversation with GiveWell, February 24, 2015 64 Rob Mather, conversation with GiveWell, February 24, 2015 65 See DRC, Idjwi South 2015 row: “Planned to follow North Idjwi. Revised timing for distribution: Q4 2015” AMF page on future distributions 66 AMF page on Nord Ubangi 2015 distribution 67 “N Ubangi, DRC Insecticide Resistance Research study pre-proposal recevied (attached) and OK’d. Proposal document in 4 weeks. Two weeks fr review. Approval if all in order: mid March 730,000 nets ETA N Ubangi Jul15” Rob Mather, email to GiveWell, February 24, 2015 68 “Malawi will be carrying out a nationwide LLIN distribution in 2015 and may need additional LLINs.84 AMF is considering funding 600,000 to 1 million LLINs85 for Concern Universal to distribute in this additional district. This would likely cost $2.4-4.0 million,86 including non-net costs (we assume AMF will fund non-net costs as it has for recent Concern Universal distributions).” GiveWell 2014 end-of-the-year review of AMF 69 We have not yet asked AMF why Malawi requested funding for 2018 instead of 2015. “2018: In addition, the NMCP through the LLINs task force is grateful for the offer from the CU/AMF partnership to continue to assist in four districts beyond the 2015 distributions, for the 2018 universal coverage distributions. In view of the above the NMCP and LLINs task force is requesting AMF/CU to undertake LLINs mass distribution campaign in four district in 2018, which will again be in line with the NMCP’s schedule and policy.” Malawi NMCP invitation for 2018 distribution AMF has already agreed to do one distribution in late 2017 in Dedza; we believe that this is thought of as part of Malawi’s 2018 universal coverage campaign, and is one of the four districts in which AMF would assist Malawi. AMF page on Dedza 2017 distribution 70 Rob Mather, conversation with GiveWell, February 24, 2015 We assume this estimate includes shipping costs, as AMF has recently been paying shipping costs in its agreements with Concern Universal. For previous Malawi distributions for which AMF has paid shipping costs, see: AMF page on non-net costs 71 AMF has shared with GiveWell the names of the countries where it is engaged in discussions regarding funding nets but has asked us not to make public the names of countries until the discussions are concluded. In part this is because AMF has agreed in principle to co-funding partnerships and wishes to respect the desire of others to have their plans kept private until distributions are finalized. 72 Rob Mather, conversation with GiveWell, February 24, 2015 73 Rob Mather, conversation with GiveWell, February 24, 2015 74 Rob Mather, conversation with GiveWell, February 24, 2015 75 Rob Mather, conversation with GiveWell, February 24, 2015 76 Rob Mather, conversation with GiveWell, February 24, 2015 77 Rob Mather, conversation with GiveWell, February 24, 2015 78 Assuming that nets cost $3 each and that non-net costs are $3 per net in Country X, a 2.5 million net distribution in Country X would cost about $15 million in total. AMF currently has a maximum of about $14.2 million in uncommitted funds. See room for more funding section. 79 Rob Mather, conversation with GiveWell, February 24, 2015 80 Rob Mather, conversation with GiveWell, February 24, 2015 Rob Mather, email to GiveWell, June 19, 2015 81 Rob Mather, conversation with GiveWell, February 24, 2015 82 AMF told us that the total net gap was 18 million nets. Rob Mather, conversation with GiveWell, February 24, 2015 In one spreadsheet that AMF sent us, AMF had calculated a need of 13,861,516 nets for 2015-2017 for mass distributions in Country Y. Including net needs that will be met through mechanisms other than mass distributions, the net need for 2015-2017 is 17,563,399. @Country Y - LLINs Gap@ Another funder is planning to fund 9 million nets for Country Y, which would leave approximately a 9 million net gap. Rob Mather, conversation with GiveWell, February 24, 2015 “In Country A, for example, the total need for nets is 9 million, and Country A is typical in this regard.” GiveWell’s non-verbatim summary of a conversation with Rob Mather and Peter Sherratt on February 6, 2015 83 When AMF first approached Country Y about the possibility of funding nets, Country Y asked AMF to cover the entirety of this remaining gap (all 9 million nets), which would have cost AMF approximately $35 million. (Note: we are not sure what is included in this $35 million estimate, e.g., whether or not it includes shipping and other non-net costs.) Rob Mather, conversation with GiveWell, February 24, 2015 AMF did not have the funds on hand to cover 9 million nets, so it offered 5.5 million nets over the next three years instead. Currently, AMF is discussing the possibility of funding 2.2 million nets in 2015, 1.8 million nets in 2016, and 1.5 million nets in 2017). Rob Mather, conversation with GiveWell, February 24, 2015 84 Rob Mather, email to GiveWell, June 19, 2015 Rob Mather, conversation with GiveWell, February 24, 2015 85 Rob Mather, conversation with GiveWell, February 24, 2015 86 It is possible that an AMF-funded distribution could be implemented through the government’s health system. According to AMF, Country Y’s health system has a positive reputation. Rob Mather, conversation with GiveWell, February 24, 2015 Alternatively, AMF told us it could partner with different NGOs (some international, some local) in each district receiving nets. Rob Mather, conversation with GiveWell, February 24, 2015 AMF told us that the post-distribution check-ups that it requires would likely be managed by one organization. Rob Mather, conversation with GiveWell, February 24, 2015 87 Rob Mather, conversation with GiveWell, February 24, 2015 We mentioned this distribution in our end-of-the-year review of AMF: “AMF is considering funding 60,000 LLINs in one country and 100,000 LLINs in two other countries,89 for a total cost of about $780,000.90 AMF told us that it would be able to sign agreements for these distributions quickly if it chose to do so and that the reason it has not is because it expects to use its funds for larger distributions instead.” GiveWell 2014 end-of-the-year review of AMF 88 If the large distribution covers the same district where this smaller distribution is planned, AMF would not want to fund the smaller distribution, since there would no longer be a need for nets in that district. Rob Mather, conversation with GiveWell, February 24, 2015 89 Rob Mather, conversation with GiveWell, February 24, 2015 “It will likely agree to fund the distribution for 60,000 LLINs because the distribution would involve a pilot project to test the use of smartphone technology for tracking LLIN distributions, which AMF is interested in learning more about.91” GiveWell 2014 end-of-the-year review of AMF 90 “Update: As of April 21, 2015 AMF is in discussion with several countries regarding contributions of a few million nets in each case, and potentially more.” GiveWell’s non-verbatim summary of a conversation with Rob Mather and Peter Sherratt on February 6, 2015. We know of 7 different countries in which AMF may be pursuing distribution opportunities. Rob Mather, email to GiveWell, November 6, 2014, Rob Mather, email to GiveWell, February 6, 2015, Rob Mather, conversation with GiveWell, February 24, 2015. 91 Rob Mather, conversation with GiveWell, February 24, 2015 92 Insecticide resistance research proposal AMF sent us information on the protocols that will be used for testing resistance: “The phenotypic assays (i.e. mosquito testing) are based on WHO procedures for tube bioassays (diagnostic and dose-response) and net testing with only minor modifications. These can be downloaded from the following URLs. http://www.who.int/malaria/publications/atoz/9789241505154/en/ malaria.lshtm.ac.uk/news-events/news/whopes-guidelines-laboratory-and-field-testing-long-lasting-insecticidal-nets .” Rob Mather, email to GiveWell, February 24, 2015 93 “Baseline (coincident or preferably just preceding LLIN-distribution) and one year post- distribution assessment of mosquito species abundance and class I/II pyrethroid resistance prevalence and level, and effect of PBO on mortality (using standardized tube assays and direct bioassay testing of LLINs) in a primary focal area in the environs of Gbadolite.” Insecticide resistance research proposal 94 “Temporal assessment (at monthly or bi-monthly sampling intervals, running from the time of the baseline to one year post-distribution) of entomological parameters (abundance, species composition, human blood-feeding rate and sporozoite infection rate, and frequencies of DNA knockdown resistance (kdr) mutations) in sentinel houses located within village clusters covered by two contrasting LLIN types (standard pyrethroid only and pyrethroid + PBO) within the primary focal area.” Insecticide resistance research proposal 95 Although AMF said in an email that it expected to approve the research plan in mid-March, we believe it intended to say mid-April based on the length of times mentioned (6 weeks from late February would be mid-April): “N Ubangi, DRC - Insecticide Resistance Research study pre-proposal recevied (attached) and OK’d. Proposal document in 4 weeks. Two weeks fr review. Approval if all in order: mid March” Rob Mather, email to GiveWell, February 24, 2015 96 In mid-March AMF published a blog post claiming that it had $19.6 million available to distribute: “We are allocating US$19.6m, sufficient to cover approximately 6 million nets.” AMF blog post, March 17, 2015 97 Rob Mather, email to GiveWell, February 24, 2015 said that AMF had about $18 million in total funds, of which $12.6 million was uncommitted (implying commitments of $5.4 million). Our understanding is that AMF has committed about $5.4 million to: North Idjwi Island and South Idjwi Island (2015): ~$370K Nord Ubangi (2015): ~$2 million Balaka (2015): ~$700K Ntcheu (2015): ~$1 million Dedza (late 2017): ~$1.3 million AMF All distributions page In mid-March, AMF said it had about $19.6 million in total funds (AMF blog post, March 17, 2015). With $5.4 million in commitments, it should now have about $14.2 million uncommitted. 98 “IRR/N Ubangi - $500,000” Rob Mather, email to GiveWell, February 24, 2015 99 ”In addition, AMF has been asked by the Malawi NMCP to continue to support malaria control activities in the four districts of Ntcheu, Balaka, Dedza and Dowa in 2018. This is a 1.5 million LLIN, US$6m commitment.” AMF blog post, March 17, 2015 Rob Mather, conversation with GiveWell, February 24, 2015 100 In Country Y, AMF is discussing an agreement of 5.5 million nets over three years, which would cost about $17 million in total (assuming about $3 per net). (More.) Rob Mather, conversation with GiveWell, February 24, 2015 However, AMF currently only has about $14.2 million in uncommitted funds. See calculation in footnote 97 above. Committing funding for Country Y’s 2016 and 2017 distributions would cost $9.9 million more than committing to only the 2015 distribution. 1.8 million nets (2016) + 1.5 million nets (2017) = 3.3 million nets. Assuming ~$3 per net, $3 x 3.3 million = $9.9 million. 101 More specifically, AMF has told us that its strategy for relatively smaller distributions is to estimate the number of nets it can purchase (by dividing its funds-in-hand by $3), then to divide this total number of nets by 4, and then to pursue a number of distributions that are of this one-quarter-of-the-total size. So, AMF is currently looking at several 1-2 million net distributions, although it is only investigating these opportunities, not engaging in detailed discussions, so as to not make false promises. Rob Mather, conversation with GiveWell, February 24, 2015 AMF told us that there are 6-8 countries with net gaps of 5 million nets or more for 2015-2016. Rob Mather and Peter Sherratt, conversation with GiveWell, February 6, 2015 102 Rob Mather and Peter Sherratt, conversation with GiveWell, February 6, 2015 ”AMF is a small organization, but believes that it has the staff capacity to handle more distributions. AMF acknowledges that negotiations with governments are often protracted, but says that any delays are usually as a result of NMCPs being under-resourced and very busy, doing their best to deal with many competing priorities. Once AMF receives a response, it acts quickly.” GiveWell’s non-verbatim summary of a conversation with Rob Mather and Peter Sherratt on February 6, 2015