Sun, 09 Dec 2018 Analysis

IMANI: On Drones, Ghana Health Service Boss, Dr. Nsiah-Asare is Very Confused

Dr. Nsiah-AsareDr. Nsiah-Asare

We have just been sent a clip of the popular Joy FM primetime morning show, Newsfile, in which Dr. Anthony Nsiah-Asare is heard discrediting IMANI’s positions on the Zipline drone program.

The gentleman seems completely at sea with the thrust of our arguments. To cure his confusion, we have summarised our detailed analysis in this brief note. He should do well to read carefully this time around.

  1. IMANI’s position is that the drone program should be piloted at a budget not exceeding $100,000 over a period of 6 months in one of the very few areas of Ghana where an emergency drone service can be justified on social welfare and public finance grounds. Such a location is the area around Saboba, Kpalba and Wapuli in the Northern region, where the topography, health demographics, and infrastructure situation warrants such an investment. During the rainy season, these communities are cut off from Yendi, their only vital link to the national health supply chain.
  2. After 6 months of piloting, a detailed evaluation report by independent researchers should guide any further investments and scale up.
  3. The $100,000 budget implies costs of 10 deliveries a day and a $45,000 overhead. Either Zipline or another service provider can operate this pilot, though for obvious reasons Zipline’s enthusiasm and clear political support give it advantages.
  4. It is noteworthy that after 2 years of implementation in Rwanda, the average daily delivery rate has been less than 10 drone deliveries per day, clearly showing that even in Rwanda, which is held up as the standard for this service by Dr. Nsiah-Asare, this program is being used for a very narrow set of interventions. The plan in Ghana to start with 600 deliveries (maximum capacity) a day within the life of the first contract is RIDICULOUS and must be revised.
  5. We point to the case of Tanzania, which after nearly 2 years of due diligence on this drone program decided to start with a small pilot program directed by its own Ministry of Health in the Lake Victoria program to establish proof of concept before making the decision whether to sign on to the Zipline offering (we know for a fact that they have complained incessantly about cost).
  6. The agreement being foisted on Parliament is very clear about how much Ghana will be paying. If a distribution center makes the capacity available for 100 deliveries a day, then the cost per month is $88000. The agreement is very clear that so long as the capacity is made available, the government needs to pay, whether or not there is uptake. Either Dr. Nsiah-Asare is confused or deliberately confusing the public with his claim that should we refuse to use the facility up to its available capacity we shall not be liable for payments.
  7. Therefore, should things go as planned, the base cost per month for the service shall be $352,000 (across the 4 distribution centers) for an average of 12000 deliveries per month or 18,000 deliveries at maximum capacity. The base cost over the life of the contract is thus $16.9 million if things go according to plan.
  8. However, there are several other variable costs, particularly the costs of tax exemptions. This is clearly shown in an analysis in the contract where these costs are tallied up to $47,000 in a sample bill of $181,928. There should be no quarrel that this is an expensive service. The implied additional cost per base is thus $23,500. The average imputed monthly cost per base is $111,500, and the likely total cost of the contract is $21,408,000.
  9. We are deliberately ignoring the costs of pooling the supplies and transporting them to the Zipline center before they can be distributed, as well as the costs of preparing the fields to receive the parachuted supplies when they arrive at the community health center.
  10. The likely cost per delivery is thus effectively between $37 and $27 per delivery. In our earlier note, we discounted this amount heavily by assuming no added costs and unrealistic full-capacity throughputs. Even doing this brought the average cost of delivery to $19.5.
  11. This “cost of delivery” is completely unrealistic for most health scenarios. In the case of blood, each drone is capable in one delivery to supply enough blood for one transfusion session. This means that the added cost for using drones in the average transfusion session is at least $19.5 (~98 GHS) and may be as high as $37 (185 GHS). We must regard this as a “transport premium”.
  12. The transport premium for delivering health commodities is a critical aspect of cost analysis in the health sector. IMANI’s argument is that using mobile cold vans, the problem of “emergency stockouts” in most deprived areas, the only real justification for drones, can be solved conclusively at a transport premium of only $2.
  13. A mobile cold van approach will “forward-position” refrigerated trucks, supplemented, where necessary, by motorcycles and bicycles (which our own “model transport portfolio” promotes) at vantage points to handle sub-district deliveries when the existing national distribution system fails particular facilities.
  14. We have noted that the delivery capacity of all the drones in all the four planned distribution centers can be carried in one “trotro” van. Buying 6 temperature-controlled mobile vans should be more than enough to implement a mobile hub-and-spoke system to address 95% of the emergency stockout problems that drones are meant to address.
  15. The drone network and its “trotro scale” capacity can never be extensive enough to meet more than 1% of the needs of the >4000 health facilities in the four regions where the drone distribution centers will be based or the 15 million outpatient visits recorded in those facilities. This is exactly why we must not spend ridiculous amounts of money on this service and should localise it to only extreme situations.
  16. It is important to point out that the Ghana Health Service that Dr. Nsiah-Asare heads is merely one agency in the health commodities supply chain in Ghana. There are other autonomous agencies under the Ministry of Health such as the Central Medical Stores and the Ghana National Drug Program, which actually deal with commodities distribution. The GHS does not have the direct control over commodities distribution. Dr. Nsiah-Asare’s posturing as if it does goes further to show a fundamental misapprehension about the Ghanaian health system.
  17. Dr. Nsiah-Asare’s confusion is also evident in his claim that there is a “technology transfer clause” in the agreement. THERE ISN’T. Training people to OPERATE drones is NOT the same thing as passing intellectual property on. In fact, the agreement is EMPHATIC that there shall be no transfer of intellectual property. Zipline maintains complete control over the intellectual property, including all the software (the most critical part of the solution) and the hardware (much of which it sources from other vendors).
  18. The good Nsiah-Asare is also so confused that he does not even realise that the only “transferring” happening is in the reverse – from Ghana to the Silicon Valley company. The agreement hands over Ghana’s medical supply chain data to Zipline and grants it worldwide license to do whatever it wants with the data.
  19. The Director-General of the GHS, persisting down this line of cluelessness, then accuses IMANI of claiming that each drone can carry one pint of blood. When this is manifestly untrue. Either the man has dyslexia or he is deliberately contemptuous. IMANI has been very thorough in analysing the program, and the evidence is here:
  20. Nsiah-Asare’s technical dyslexia continues when in assessing the equivalence of mobile cold vans he brings up as an analogy, patients using taxis to carry commodities from facility to facility. This is ridiculous. The idea that the alternative to using drones is every patient boarding trotro on their own to pick up blood, snake serum, or ORS from one facility to the other shows the level of analysis that must have gone into the program design under his leadership.
  21. On the cost of blood collection, screening, and storage, Nsiah-Asare’s position is that IMANI should have used the “retail” level cost at the National Blood Transfusion Service. With all due respect, that’s not how proper cost benchmarking is performed. Cost-benchmarking does not take for granted all the inefficiencies of the system when trying to establish the “true cost” of an activity. We based our analysis on several studies (such as Dosunmu et al, 2017; van Hulst, 2010 etc.) in order to come up with rigorous benchmark costs for screening and cold-storage in a predominantly “voluntary blood donation system”, which Ghana aspires to and Rwanda has worked hard to attain.
  22. At any rate, let us not allow the gentleman to confuse the debate. At between $11.5/kg to $21/kg, drone health commodities transportation is extremely expensive! A simple intuitive way to understand this is to look at courier prices. For example, the upper bound estimate for the proposed drone service is higher than the price per kg for DHL deliveries between Ghana and London and Ghana and Frankfurt.
  23. Of course, scale dynamics plays a major part in the analysis. Which is of course the whole point. The very nature of drone transport makes economies of scale impossible in the Ghana of today, and should therefore see drone transport restricted to emergency deliveries in a very small corner of Ghana, for which reason an initiative of this type in a nation like ours should start at $100,000 and NOT $20 million plus!
  24. Nsiah-Asare disputed our claim that over the last 2 years, Rwanda’s delivery rate has averaged about 6 per day. Zipline’s own PR is amply clear about these figures (example: this article by CNBC on which the company clearly collaborated with the writers: The company’s 5000 delivery-flights over the timeline of service in Rwanda at the time of the story works out to 6.7 deliveries per day on average. Clearly the gentleman did no serious research when he went to Rwanda.
  25. The total cost of the Zipline program in Rwanda - which was on a "per delivery" basis rather than the "capacity charge" model being used in the Ghana case - to the Government of Rwanda has been less than $150,000 in the two years that the program launched. Compare that to Ghana where, for reasons known only to Nsiah Asare and his per diem collecting bureaucrats, the program shall cost more than $10 million over a similar span of 2 years! Rwanda has "targeted" the project surgically. We unfortunately have not.
  26. Having listened to his description of the upcoming rollout of the first distribution center at Suhum, to cover a 160 kilometer diameter, we are quite alarmed. He does not appear to see the absurdity of a scheme that involves first pooling scarce health commodities at the Regional Medical Stores in Koforidua, then transporting them to a private facility in Suhum so that they can be catapulted in streams to facilities in the Afram Plains.
  27. Consider that this is a country where the Central Medical Stores itself lack refrigerated trucks to support transportation to regions. And where regional medical stores have no temperature-controlled trucks to cover even 5% of districts. Last time we checked there were less than 15 properly equipped, HAACP-compliant, distribution trucks for the whole country. If we have $20 million to spend to improve the delivery of medical commodities, that is where we should start. In fact, any emergency system, including the drone-based one, requires that we sort out the road distribution system FIRST since the commodities must first be pooled and transported to the Zipline distribution centers!
  28. One of the strongest indications that Dr. Nsiah-Asare is completely ignorant about the subject he has been droning on about is the claim that there are no other unmanned aerial health commodity delivery programs around the WORLD apart from Zipline’s! This is pathetic! There are several pilots ongoing in the world today to shape drone delivery interventions that actually make sense, since drones are merely complementary to a sound health supply chain network.
  29. Here are obvious examples: a) the Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) program in Tanzania mentioned earlier; b) right here in Ghana, the Dutch Government and the UNFPA rolled out a reproductive commodities delivery program by drones that has failed to scale because Government refuses to fund it (clearly, no per diem opportunities for GHS bureaucrats to travel out; read more about this here:; c) Puerto Rico’s Skypod program; various initiatives by Flytrex and Matternet (eg. Jointly with Matternet in Papua New Guinea); and d) the Vayu initiative in Madagascar. If this was merely about the urgency of drone transport, why didn't Dr. Nsiah Asare hop from radio station to radio station asking for funds for the earlier 2016 UNFPA drone program in Ghana that became stagnant due to lack of funds?
  30. If a landscape analysis led by Dr. Nsiah-Asare could not surface all these many initiatives to learn lessons and understand the importance of careful piloting, then he is clearly much too myopic to be allowed to run amok purporting to be directing this exercise.
  31. Given how porous his awareness about critical issues is, Dr. Nsiah-Asare does not inspire our confidence in this project. Hopefully, a technical panel composed of actors from other critical agencies beyond the GHS shall be put together to implement a small pilot in the Saboba area or one of the other few places that the GHS’s own analysis of the health transport system show cannot be adequately served by vans, bicycles and motorcycles especially during the rainy season. It would appear that Nsiah-Asare may not even be aware of the “model transport portfolio” guiding the very agency he heads.

129201815158 vaqduhgtsn 20181209 135037

129201815158 1h830n4ayu 20181209 135140