1 00:00:04,370 --> 00:00:07,315 Alex Ergo: Hi everyone, welcome back to Health Systems Pathways. 2 00:00:07,860 --> 00:00:17,980 I'm Alex Ergo, PSI's Director of Health Systems, and I'm excited to bring you the third live episode recorded at the Health Systems Research Symposium in Nagasaki. 3 00:00:18,283 --> 00:00:23,863 In today's episode, we focus on primary health care and the critical role of health system design. 4 00:00:24,283 --> 00:00:35,903 We'll be hearing about how better resource allocation and smarter data use can optimize service delivery, and how we can get better at performance tracking to drive improvements in health outcomes. 5 00:00:36,898 --> 00:00:40,258 It's an inspiring conversation packed with practical insights. 6 00:00:40,268 --> 00:00:41,678 So, let's dive in. 7 00:00:46,035 --> 00:00:46,565 Hi, everyone. 8 00:00:46,785 --> 00:00:49,845 We're back here and with me is Nirmala Ravishankar. 9 00:00:49,975 --> 00:00:51,015 Nice to have you with us. 10 00:00:51,355 --> 00:00:54,375 Nirmala, would you like to start by briefly introducing yourself? 11 00:00:54,615 --> 00:00:54,665 Nirmala Ravishankar: Sure. 12 00:00:54,905 --> 00:00:55,625 Thanks, Alex. 13 00:00:55,625 --> 00:01:02,650 My name is Nirmala and I'm deputy director of the primary health care team at the Gates Foundation. 14 00:01:03,030 --> 00:01:05,250 And yeah, I'm excited to meet with you today. 15 00:01:05,449 --> 00:01:05,809 Alex Ergo: Great. 16 00:01:05,809 --> 00:01:11,889 So, let me start by asking you, what are some of the priorities that you're focusing on with your new role? 17 00:01:12,525 --> 00:01:12,765 Nirmala Ravishankar: Yeah. 18 00:01:12,765 --> 00:01:47,565 We have a primary health care strategy and it's got a few different moving parts, but the part that I've been thinking a lot about and that perhaps we could start with is a question around health system design, or that's the term we've been using to think about it, which is to say, especially for public sector delivery, but certainly for for delivery of primary health care across sectors, how and where in government, are people thinking about whether the existing facilities are in the right places? 19 00:01:47,885 --> 00:01:54,015 Whether our health workers are deployed in the most optimal manner? 20 00:01:54,415 --> 00:01:59,245 And are there ways to improve that allocation or reallocate? 21 00:01:59,625 --> 00:02:04,745 Such that you're optimizing for output and indeed outcomes? 22 00:02:05,155 --> 00:02:09,395 And what kind of data you use to make those kinds of decisions, right? 23 00:02:09,395 --> 00:02:18,965 Because it's not an abstract question, but rather being able to say, 'Hey, here's where the people are and here's where the needs are'. 24 00:02:19,315 --> 00:02:30,885 And compared to the needs, we don't have sufficient number of facilities and we have the right facilities, but we don't have the right people where those facilities are. 25 00:02:31,205 --> 00:02:43,919 So it's an optimization problem for which you need data, you need as well operational data to see what's the readiness of the existing facilities, you need output data. 26 00:02:44,099 --> 00:02:59,309 So it's a planning question first and foremost, but that quickly leads you into thinking about what kind of data is the health system currently producing, and is it sufficient to answer these kinds of questions? 27 00:02:59,659 --> 00:03:32,888 And there, I think we're feeling our senses that most countries, low and middle-income countries, actually have a lot of existing data systems, but they're not often connected in a way that you can look at all sides of the equation, by which I mean you can start with how much money is going into the system to buy what kind of inputs; the workers, the commodities, et cetera, and then leading to what kinds of, or what amount of outputs. 28 00:03:33,178 --> 00:03:52,633 Because if you could line up those three pieces, then you could really explore, if you think of a bunch of facilities or a bunch of districts, why are some facilities or some districts doing better than others, and you can start making those comparisons if you could link the data. 29 00:03:53,043 --> 00:04:02,893 But right now, in many places where we're working, you have public financial management systems that track the money, especially public money. 30 00:04:03,463 --> 00:04:14,223 You have the district health management information systems that track output; how many services did we deliver today or in the past month? 31 00:04:14,663 --> 00:04:33,082 And then you have other health information systems that are for health workers and a separate system for supply chain and another system that will be just for some vertical disease program, and that interoperability between those systems remains quite weak. 32 00:04:33,292 --> 00:05:07,727 So, just to then go back to your question, we're both thinking about how to strengthen capacities and the muscle within the system to ask these questions about whether existing resources, and by resources, money, but also people, are being allocated in a way to optimize output and linking or informing those kinds of decisions with good, robust, timely and high quality data about flow of funds, about by 33 00:05:10,037 --> 00:05:10,737 operations, 34 00:05:13,517 --> 00:05:16,773 about facility operations , are there drugs on the shelves? 35 00:05:16,773 --> 00:05:17,180 Are the health-workers showing up to work? 36 00:05:17,180 --> 00:05:17,337 And then output. 37 00:05:17,337 --> 00:05:23,277 So that's in a nutshell, maybe a large nutshell, what we're thinking about and working on. 38 00:05:23,787 --> 00:05:24,467 Alex Ergo: That's excellent. 39 00:05:24,487 --> 00:05:27,087 Yes, I can see that very often. 40 00:05:27,087 --> 00:05:32,127 It's still very much top down and based on norms rather than on the reality on the ground. 41 00:05:32,127 --> 00:05:40,987 And so, what you just described is, I would imagine, the first step once you have your resources allocated to the right place, that's a good start. 42 00:05:41,027 --> 00:05:42,407 But then how do you. 43 00:05:42,862 --> 00:05:51,002 Track performance and make sure that what's being done in those delivery points meets the needs of the patient or the client. 44 00:05:51,292 --> 00:05:51,992 Nirmala Ravishankar: Absolutely. 45 00:05:52,012 --> 00:06:00,372 So, you've hit the nail on the head in terms of the need for tracking performance and holding different system actors accountable. 46 00:06:00,448 --> 00:06:01,579 And so, it's a loop, right? 47 00:06:01,599 --> 00:06:18,874 It's a virtuous cycle between starting with good planning, which needs to be informed by robust data, and then on the flip side, you're holding people accountable for performance against a certain plan that's articulating what the vision is. 48 00:06:19,134 --> 00:06:24,474 So, they go hand in glove, these two pieces, the realistic planning, as well as the performance management. 49 00:06:24,614 --> 00:06:25,444 Alex Ergo: Yeah, excellent. 50 00:06:25,544 --> 00:06:33,144 And do you see already good examples where, countries are thinking that way and actually putting it into practice? 51 00:06:33,224 --> 00:06:33,504 Nirmala Ravishankar: Yeah. 52 00:06:33,594 --> 00:06:59,779 So in Pakistan one of my colleagues is working closely with a couple of provincial governments there as well as implementing partners, our partners to think about a low cost and quick, in real time, an approach for tracking service readiness and tracking outputs and then feeding that. 53 00:07:00,274 --> 00:07:02,014 into planning. 54 00:07:02,214 --> 00:07:13,314 Because if you just think about it, if you're doing the performance tracking and you say, wait a minute, this facility is continuously lagging, then you can start exploring why that is. 55 00:07:13,314 --> 00:07:16,874 And maybe some of those issues are within the control of the facility. 56 00:07:16,874 --> 00:07:21,684 But perhaps there are other issues that go that are more systemic. 57 00:07:22,014 --> 00:07:31,164 And then you need to build that into your plans as a local district or as the provincial government or state government in a country. 58 00:07:31,334 --> 00:07:45,645 So, we've seen that kind of feedback loop coming into play, even though our entry point into that work was more from the performance management end of things, because that's where government was most keen to start. 59 00:07:45,885 --> 00:07:52,270 But you can start seeing how it's influencing the planning processes at different levels of government. 60 00:07:52,511 --> 00:07:57,651 Alex Ergo: Are you referring to these rapid cycles through the faster approach, or is that different? 61 00:07:57,831 --> 00:07:58,611 Nirmala Ravishankar: That's different. 62 00:07:58,621 --> 00:08:00,301 That's also something we're working on. 63 00:08:00,301 --> 00:08:08,901 But you can see how that's that's also an attempt to very quickly understand the performance at facility level. 64 00:08:10,221 --> 00:08:16,212 Ultimately, when you think about what information you may want to track in a system, A.P.H.C. 65 00:08:16,212 --> 00:08:17,622 service delivery system. 66 00:08:17,715 --> 00:08:25,735 You need the administrative data, the data that the facilities are tracking every day and reporting into the system. 67 00:08:26,025 --> 00:08:31,005 But you also need some way to independently check the veracity of that. 68 00:08:31,050 --> 00:08:44,932 Right now, we do surveys, but if you do a large survey of all facilities, that's a costly and time consuming enterprise, which you're not going to be able to do every month or even every year. 69 00:08:45,332 --> 00:08:51,432 And so, what are some ways to rapidly check for service readiness? 70 00:08:51,732 --> 00:08:55,412 It's something that we've been thinking about and making investments around. 71 00:08:55,692 --> 00:08:58,312 And then you can think of a few additional pieces, right? 72 00:08:58,312 --> 00:09:06,152 One is just the facility level readiness and tracking that both through administrative data and independent monitoring. 73 00:09:06,412 --> 00:09:16,502 But you also want to understand the patient experience and people could go to the facility and get the services, but if they're not treated with respect, if the waiting times are really long... 74 00:09:16,762 --> 00:09:20,432 so you need to also track the patient experience. 75 00:09:20,752 --> 00:09:27,812 And then the final piece is what is not covered by all of this, are the people who are not going to get care. 76 00:09:28,092 --> 00:09:33,392 You wouldn't be capturing them from either of those other two, other two exercises. 77 00:09:33,432 --> 00:09:40,912 So you also need population level surveys to understand why some people are still not going to seek care when they need it. 78 00:09:41,102 --> 00:09:42,332 What are the barriers? 79 00:09:42,352 --> 00:09:43,642 Is it financial? 80 00:09:43,832 --> 00:09:45,092 Is it physical access? 81 00:09:45,092 --> 00:09:45,942 Is it knowledge? 82 00:09:45,952 --> 00:09:47,122 Is it something else? 83 00:09:47,132 --> 00:10:01,197 So, these are all the different components that you need to really truly be able to understand the current performance of the system as well as to design for a better service delivery system. 84 00:10:01,595 --> 00:10:03,405 Obviously, you're not going to get there overnight. 85 00:10:03,425 --> 00:10:27,205 You're not going to build out all these capacities overnight, but starting somewhere and looking for those virtuous cycles if it's from performance management to realistic planning, great, but in some other country, maybe the opportunity is first to start with a big service delivery reform that state government or national government is enthusiastic about. 86 00:10:27,345 --> 00:10:43,395 And so whatever that opportunity may be, that opening might be to start there and then build towards this comprehensive information system to inform realistic planning and performance management for primary health-care. 87 00:10:44,029 --> 00:10:45,292 Alex Ergo: This is super exciting. 88 00:10:45,452 --> 00:10:48,142 And yes, this is not going to happen overnight. 89 00:10:48,202 --> 00:10:52,062 But at least you have some sort of North Star that you can follow. 90 00:10:52,432 --> 00:10:55,552 And do you see these pieces already in the making? 91 00:10:55,575 --> 00:10:57,845 For example, you gave the example of Pakistan. 92 00:10:57,855 --> 00:11:04,015 Do you see all these different pieces already being worked on separately and with the idea that they will come together? 93 00:11:04,015 --> 00:11:05,395 Or is that the next step? 94 00:11:06,095 --> 00:11:11,585 Nirmala Ravishankar: I would say that, in some places, these pieces are starting to come together really well. 95 00:11:12,065 --> 00:11:30,805 But I think this kind of more comprehensive approach is something that our team has been working on for a few years, but we're getting clearer about what we mean and clearer about what kind of investments we need to make to catalyze this kind of change. 96 00:11:31,065 --> 00:11:34,805 I think we're also looking at many of these changes. 97 00:11:35,260 --> 00:11:39,340 We're starting with some subnational governments, right? 98 00:11:39,340 --> 00:11:46,820 So, primary health care delivery is by and large in many big countries controlled by local governments. 99 00:11:47,070 --> 00:11:52,010 And so if you enter a country with 50 local governments, you're not working with all of them. 100 00:11:52,260 --> 00:12:22,813 So starting with one or two where the opportunities present themselves and almost arriving at a proof of concept and then taking learnings from that work to advocate for and push for change in that country first and foremost because especially you know learning between states within a country is easier than learning between countries because you've already controlled for many things, political system, culture, by and large. 101 00:12:23,003 --> 00:12:42,073 And creating those opportunities for diffusion and learning within countries from subnational work in one or two targeted places and then, of course, scaling it even broader than that to create opportunities for regional and global knowledge sharing is also something that we invest in heavily. 102 00:12:42,593 --> 00:12:43,883 Alex Ergo: This is a great agenda. 103 00:12:43,893 --> 00:12:46,833 I know really something exciting to focus on. 104 00:12:46,833 --> 00:12:48,143 So, thank you for sharing that. 105 00:12:48,163 --> 00:12:52,273 Maybe before we end, a last question about the conference. 106 00:12:52,403 --> 00:12:54,383 This is our third day into the conference. 107 00:12:54,403 --> 00:12:56,323 What are some of the takeaways? 108 00:12:56,353 --> 00:12:57,993 What surprised you? 109 00:12:57,993 --> 00:13:00,383 Or what are some of your observations? 110 00:13:00,543 --> 00:13:15,504 Nirmala Ravishankar: I think closer to home for me, you know, I was one of the organizers of a session on Monday, that was all about big health systems reforms and what motivates countries to undertake health systems reforms. 111 00:13:15,836 --> 00:13:20,684 And the big message there was to start with a clear articulation of the problem. 112 00:13:21,014 --> 00:13:43,274 And it sounds so simple, it sounds intuitive and obvious, but there are numerous examples of countries starting with a solution rather than the problem and to be perfectly blunt, sometimes conferences like this can lead to there's this term Lant Pritchett, I think, coined about isomorphic mimicry, right? 113 00:13:43,274 --> 00:13:47,734 You go to a conference, you hear about, oh, this country has done X and that sounds really great. 114 00:13:47,974 --> 00:13:54,594 And so, we should also come and do X, but does X actually respond to the problem in your country? 115 00:13:54,594 --> 00:14:04,714 So, starting with a very clear articulation of the problem you're trying to solve for and the objective that you're trying to achieve and then back into the right solution. 116 00:14:05,024 --> 00:14:30,341 So that was that was a message from our session, but big picture, this conference is shining a bright bright light on the need to think about climate change, and you know what we in the health sector folks working on health systems can do about climate change as well as how the health system can be made to be more resilient to climate change. 117 00:14:30,341 --> 00:14:33,331 And I think that's a really timely and important conversation. 118 00:14:33,611 --> 00:14:39,816 It is not one that leads to easy answers but one that we need to really grapple with. 119 00:14:41,376 --> 00:14:55,086 Alex Ergo: Yeah, and I would say that all the things that you've discussed in the beginning are actually contributing to make the health system more resilient because you can become more agile and responsive to where the needs are, so in some way it comes together. 120 00:14:55,296 --> 00:14:56,106 Nirmala Ravishankar: Absolutely. 121 00:15:00,198 --> 00:15:09,258 Alex Ergo: Thank you to Nirmala for sharing her insights on how better planning, performance management and data use can transform primary health care. 122 00:15:09,878 --> 00:15:21,598 Today's episode highlighted the importance of rethinking health system design, whether it's optimizing resource allocation, strengthening data systems or creating feedback loops to improve performance. 123 00:15:21,891 --> 00:15:24,501 Thank you for tuning into Health Systems Pathways. 124 00:15:24,841 --> 00:15:30,451 If you enjoyed this episode, don't forget to subscribe or follow us wherever you get your podcasts. 125 00:15:30,741 --> 00:15:37,031 We'll bring you more fresh insights from the halls of the HSR Symposium in Nagasaki very soon. 126 00:15:37,501 --> 00:15:39,641 Until next time, I'm Alex Ergo. 127 00:15:39,731 --> 00:15:43,831 Let's keep working together to build stronger health systems that work for people.