Reflections on the use of the World Health Organization’s (WHO) OneHealth Tool: Implications for health planning in low and middle income countries (LMICs)

The World Health Organization (WHO) launched the OneHealth Tool (OHT) to help low and middle income countries to develop their capacities for sector-wide priority setting. In 2016, we sought to use the OHT to aid the Philippine Health Insurance Corporation (PHIC), the national health insurer of the Philippines, in decisions to expand benefit packages using cost-effectiveness analyses. With technical support from the WHO, we convened health planning officers from the Philippine Department of Health (DOH) and the Philippine Health Insurance Corporation (PHIC) conduct generalized cost-effective analyses (GCEA) of selected un-financed noncommunicable disease interventions using OHT. We collected epidemiological and cost data through health facility surveys, review of literature such as cost libraries and clinical practice guidelines, and expert consultations. Although we were unable to use GCEA results directly to set policy, we learnt important policy lessons which we outline here that might help inform other countries looking to inform service coverage decisions. Additionally, the entire process and GCEA visualizations helped high-level policymakers in the health sector, who have traditionally relied on ad hoc decision making, to realize the need for a systematic and transparent priority-setting process that can continuously provide the evidence needed to inform service coverage decisions.

Striving for universal health coverage requires priority setting to maximize health gains from limited resources (Chalkidou et al., 2016). Countries like the Philippines struggle with building technical and institutional capacity for health technology assessment (HTA).
In 2015, the national insurer Philippine Health Insurance Corporation (PHIC) wanted to redefine their health benefits package. We developed a list of priority conditions based on disease burden and a corresponding list of cost-effective interventions (Wong et al., 2018). This paper documents our subsequent attempt on a more definitive list of interventions using cost-effectiveness analyses (CEA) in the Philippine setting. At that time, however, we had insufficient technical capacity to conduct the volume of necessary primary CEAs within the timeframe of our policy window. We sought the help of the World Health Organization (WHO) to conduct generalized costeffectiveness analysis (GCEA) using their OneHealth Tool (OHT) software.
The appeal of GCEA and OHT was its perceived feasibility for our purposes. OHT had been used successfully in 20 low and middle income countries (LMICs) for sector-wide planning (Avenir Health, 2017), and WHO-CHOICE had conducted regional-level GCEA for a good range of diseases in our priority list of conditions (World Health Organization, 2017). Because GCEA uses a "do nothing" comparator, we would be able to simultaneously analyze multiple diseases and new and existing alternatives within and outside the health sector to determine the most efficient mix of interventions to fund at the national-level (Hutubessy et al., 2003). Lastly, GCEA could be done through the user-friendly OHT software with its built-in health impact models, templates for costing and epidemiological inputs, and data visualization functions.
We hope that our reflections on our experience with GCEA using OHT will aid other LMICs looking to scale up their priority setting efforts.

Methods
Based on the available models in the OHT at that time (April 2016), twelve interventions across three diseases -cardiovascular disease, diabetes, chronic obstructive pulmonary disorder -were chosen for analysis. OHT has global or regional defaults for all of the values in the models, and we attempted to replace all of the defaults with the best available local data. To do this, we collected epidemiological and cost data from health facility surveys, desk reviews, and clinical experts. A summary of data sources for each input is detailed in Table 1.
We invited WHO-Geneva to conduct a five-day workshop on the use of OHT, with DOH and PHIC planning officers, and some senior officials as participants. Over the course of the workshop, we used the default parameters of the model as a starting point of discussion. We used a consensus approach in modifying these defaults. We first consulted available data we had gathered, then all workshop participants discussed iteratively which data inputs and resource requirements were appropriate in the Philippine context until a consensus was reached.
On the last day, we plotted the cost-effectiveness results on an isoquant graph. Each intervention represents a point, and the graph is to be interpreted diagonally. The graph ( Figure 1) shows the cost of the intervention on the horizontal axis, and effectiveness on the vertical axis. We have also added price tags to each data point to indicate the five-year budget impact.

Lessons learned
Our study raised awareness among policy makers that cost-effectiveness can be used for making decisions on service coverage. They found the GCEA results straightforward and understandable. However, for the policy question, GCEA methods were deemed insufficient to satisfy policy makers' demands. GCEA was geared for choosing a complement of mutually compatible interventions based on a 'do-nothing' scenario.
The results ended up validating the programs already covered, but failed to address the question of expanding service coverage. We realized that a CEA with current practice as the comparator would be more appropriate.
Another limitation of GCEA is the limited availability of models, making it difficult to conduct a sector-wide exercise that completely incorporates the Philippine baseline context. While more models have been added since our last use, such as modules on neglected tropical diseases, it will take time before the full breadth of interventions will be covered. Most of the disease packages available were based on WHO's list of global priorities, which were already covered in the Philippines. In addition, the process of modifying defaults in the OHT was more difficult than expected. Expert opinion and clinical practice guidelines varied widely, which made it challenging to reconcile the input values, especially in the absence of a local methods guideline for economic evaluation.
The undertaking also led to the policy makers' realization that no systematic process existed within PHIC for developing

Amendments from Version 1
We have added one more person in the Acknowledgments section, who was forgotten from the version 1. We apologize for this oversight. No other changes have been made. benefit packages. The failure in providing an answer to the question of "Which intervention to cover next?" through the GCEA pushed all stakeholders to consider setting up a process and value framework for priority setting alongside an institution that will facilitate such processes. While PHIC and DOH are not completely new to HTA, there was a need to establish a single, harmonized priority setting process for the health sector that would enable these two institutions to make better policy decisions based the questions they are currently faced with. This envisioned an HTA system, characterized by societal principles, with specific criteria and steps and supporting legal framework that would respond to the next level of questions such as which drug to reimburse, given that both institutions are flooded by demands from various stakeholders to cover a wide range of interventions.

REVISED
This whole exercise re-emphasizes that defining a health benefits package cannot be a one-off exercise. It is influenced by a myriad of factors, and entails much stakeholder consultation and awareness of societal and institutional values. We recognize that the next step for the Philippines is to embed in the health system a process of both evidence generation and use for making investment decisions in the public sector. Developing capacity at both technical and sectoral dimensions must also be prioritized. Despite the limitations of the GCEA, it was a practical exercise to demonstrate the importance of evidence use in supporting policy surrounding service coverage decisions.

Data and software availability
The OneHealth Tool (OHT) may be downloaded freely at from Avenir Health. At the time this study was conducted, the Spectrum version of the OHT was 5.42. The results may be recreated by following the detailed documentation of changes to the OHT defaults found in Supplementary File 1.
Data entered into OHT was obtained from sources listed in Table 1 Competing interests No competing interests were disclosed.

Grant information
The project was funded jointly by the United Nations Children's Fund (UNICEF) and the Philippine Health Insurance Corporation (PHIC).
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. They clearly describe the rationale and aim of using this tool to support redefining the benefits package of the national insurer of the Philippines. Most importantly, they identify key limitations on various levels: 1) limitations of the OHT itself; 2) limitations in availability of models and data specific to the Philippine context; and 3) the unmet need to establish a single, harmonized priority setting process for the health sector that would enable multiple stakeholders (in this case the Department of Health and the Philippine Health Insurance Corporation) to make better policy decisions based on the questions they are currently faced with. Based on these limitations and their experiences the authors provide general recommendations that are relevant to the Philippines and to other LMICs.

Open Peer Review
The paper is valuable and sound, given that its aims are quite humbly to share reflections on the use of the OHT for a priority setting exercise. However, these reflections -however well described and sensibleshould not be read as (and are not intended to be) the result of a systematic evaluation of experiences based on all stakeholders inputs. A degree of subjectivity should be acknowledged and the paper would have benefit from more specific or detailed suggestions to improve the priority setting exercise and experience, from the perspective of the different participants. That said, this paper is a valuable contribution to the literature and more examples such as these should be welcomed and shared to drive the HTA field forward, globally as well as locally.

Are sufficient details of methods and analysis provided to allow replication by others? Partly
If applicable, is the statistical analysis and its interpretation appropriate?

Is the study design appropriate and is the work technically sound? Yes
Are sufficient details of methods and analysis provided to allow replication by others? Partly If applicable, is the statistical analysis and its interpretation appropriate? Not applicable Are all the source data underlying the results available to ensure full reproducibility? Partly

Are the conclusions drawn adequately supported by the results? Yes
No competing interests were disclosed.

Competing Interests:
I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
The benefits of publishing with F1000Research: Your article is published within days, with no editorial bias You can publish traditional articles, null/negative results, case reports, data notes and more The peer review process is transparent and collaborative Your article is indexed in PubMed after passing peer review Dedicated customer support at every stage For pre-submission enquiries, contact research@f1000.com