Childhood Leukemia Survival in the US-Mexico Border: Building Sustainable Leukemia Care Using Health Systems Strengthening Models

PURPOSE Pediatric leukemia outcomes are poor in most low- and middle-income countries (LMICs) and exacerbated by health care systems ill equipped to manage cancer. Effective leukemia management in LMICs involves curating epidemiologic data; providing health care workforce specialty training; developing evidence-based treatments and supportive care programs; safeguarding access to medications and equipment; providing patient and family psychosocial, financial, and nutritional support; partnering with nongovernmental organizations, and ensuring treatment adherence. METHODS In 2013, through a partnership between North-American and Mexican institutions, we used the WHO Framework for Action, a health systems strengthening model to implement a leukemia care sustainable program aimed at improving acute lymphoblastic leukemia (ALL) outcomes at a public hospital in Mexico. We prospectively assessed clinical features, risk classification, and survival outcomes in children with ALL at Hospital General-Tijuana from 2008 to 2012 (preimplementation) and from 2013 to 2017 (postimplementation). We also evaluated program sustainability indicators. RESULTS Our approach led to a fully-staffed leukemia service, sustainable training programs, evidence-based and data-driven projects to improve clinical outcomes, and funding for medications, supplies, and personnel through local partnerships. Preimplementation and postimplementation 5-year overall survival for the entire cohort of children with ALL, children with standard-risk ALL, and children with high-risk ALL improved from 59% to 65% (P = .023), 73% to 100% (P < .001), and 48% to 55% (P = .031), respectively. All sustainability indicators improved between 2013 and 2017. CONCLUSION Using the health systems strengthening WHO Framework for Action model, we improved leukemia care and survival in a public hospital in Mexico across the US-Mexico border. We provide a model for the development of similar programs in LMICs to sustainably improve leukemia and other cancer outcomes.


INTRODUCTION
Despite significant strides in curative treatments made over the past 4 decades, acute lymphoblastic leukemia (ALL) remains one of the leading causes of childhood death and morbidity after infancy. 1 In high-income countries (HICs), 5-year overall survival (OS) ranges from 80% to 90%.This dramatic improvement has been attributed to advancements in diagnostic technologies, risk-stratification systems, and treatment protocols driven by cooperative clinical trials. 2,3evertheless, ALL survival outcomes have not been equitable around the world.As the global pediatric cancer burden grows in low-and middle-income countries (LMICs), [4][5][6] ALL OS in LMICs has been reported between 10% and 60%. 7,8hese disparate outcomes could be mitigated with wider availability of critical, specialized, and supportive care. 5,6,91][12] State of the art diagnostic and clinical services, access to pediatric oncologists, and specialized and critical care training for all providers are limited in many Mexican institutions. 13To mitigate these disparities in access and outcomes, we established in 2008 a twinning program between Rady Children's Hospital San Diego, St Jude Children's Research Hospital, and Hospital General-Tijuana (HGT), Mexico, in the US-Mexico border region. 6,14Although this model improved access to care and clinical outcomes at HGT from 2008 to 2013, its dependency on the institutions in HIC required an alternative strategy to ensure long-term sustainability and independence.To address this, in 2013, we used the WHO health systems strengthening Framework for Action model 15 to implement measures to decrease the dependency of the twinning program on external resources, promote sustainability, and continue to improve ALL outcomes in the public sector in Mexico.We describe the developmental milestones on the basis of the WHO Framework for Action and its early impact on clinical outcomes for patients with ALL at HGT preimplementation from 2008 to 2012 and postimplementation from 2013 to 2017.We also report indicators of sustainability in 2013 and 2017.

Context and Health Systems Strengthening Model for Developing High-quality Leukemia Care
HGT is a major public referral hospital in northwestern Mexico, serving approximately three million people.Yet, in 2008, the hospital had limited treatment options, and outcomes for children with cancer were dismal in Tijuana and the state of Baja California.HGT had neither a dedicated pediatric oncology unit nor a pediatric intensive care unit.There were no pediatric oncologists or pediatric oncology nurses.Diagnostic equipment was severely limited, and stock outs of medications were frequent.Responding to the need for high-quality pediatric cancer care in the US-Mexico border region, Rady Children's Hospital San Diego in partnership with St Jude Children's Research Hospital launched a collaborating twinning program at HGT in 2008. 6,14Twinning involves a partnership between pediatric cancer centers in HICs and LMICs aimed at improving survival and other clinical outcomes by sharing knowledge, expertise, and resources.
As the number of patients with ALL increased and the needs escalated, we adopted the health systems strengthening WHO Framework for Action 15 model in 2013 as a development roadmap to implement an independent and sustainable leukemia care program.The building blocks of the WHO Framework for Action describe essential and interdependent functions necessary for the optimal performance of a health system and include service delivery, workforce, information systems, access to essential medicines, financing, and local leadership/governance.We used the WHO Framework for Action as a guide in setting substantive and comprehensive milestones necessary for developing sustainable capacity for high-quality pediatric leukemia services (Fig 1).

Leukemia Care Program Implementation: Application of the WHO Framework for Action Model
We conducted a needs assessment by using an instrument adapted from St Jude Children's Research Hospital's needs assessment tools 14 and developed and set a 10-year action plan with objectives outlined according to the WHO Framework for Action.

Block 1. Health Service Delivery
To ensure the delivery of effective, safe, and high-quality clinical care with minimum waste, we focused on four main areas.
Infrastructure.Our first objective was to increase capacity through pediatric leukemia-specific infrastructure within

CONTEXT Key Objective
To describe the application of the health systems strengthening WHO Framework for Action to a twinning program between Rady Children's Hospital San Diego, USA, and Hospital General-Tijuana, Mexico, to evaluate its impact on capacity building for high-quality, sustainable care and on improving survival for children with leukemia in a low-and middle-income country (LMIC) setting.

Knowledge Generated
Our initiative resulted in a fully staffed pediatric leukemia service with protocols, organization, and structure designed to address each of the six building blocks of the WHO Framework for Action model.Preimplementation and postimplementation results demonstrate significant improvements in survival in children with leukemia in Baja California, Mexico, and improved program sustainability indicators.

Relevance
The success of the application of the WHO Framework for Action model to our twinning program supports the incorporation of the six building blocks when developing pediatric leukemia and other cancer control programs in LMICs.
the dedicated Pediatric Hematology Oncology Unit (PHOU) inaugurated in 2008 at HGT. Clinical monitoring and supportive care.To enhance the ability to monitor patient status and respond to clinical needs, we introduced the Pediatric Early Warning Score (PEWS) system. 16We secured a Mexican federal grant, developed the PEWS team, and trained all PHOU staff according to St Jude Children's Research Hospital PEWS implementation guidelines.Additionally, the Golden Hour project was established to ensure faster antibiotic initiation during neutropenic fever events to reduce sepsis risk. 17,18HGT staff developed a systematized protocol for supply inventory, indication, and administration of antibiotics.

ALL treatment. Led through bimonthly meetings with a
Spanish-speaking pediatric hematologist/oncologist (P.A.) with expertise in pediatric leukemia from Rady Children's Hospital San Diego, our team initiated an action plan to establish HGT's first locally adapted chemotherapy protocol 19 and guidelines for blood products transfusions adapted to resource-limited settings, created training materials and milestones for assessing the impact, and established a weekly leukemia board to identify areas of improvement related to management of ALL.We also began the Project MEJOR (better in English), a protocol for applying a comprehensive and system-by-system treatment plan for each patient with ALL in the PHOU during daily rounds.
Treatment adherence.To address treatment abandonment, we initiated projects to improve adherence to treatment, particularly during the 1-to 2-year long ALL maintenance chemotherapy.Nurses led education initiatives for all patients with ALL and their caregivers.

Block 2. Health Workforce
We focused our training and education efforts on diagnostics and specialized outpatient and inpatient care for critically ill patients with ALL.We appointed a lead pediatrician and two nurse educators at HGT to develop a primarily on-site training in-person curriculum in pediatric leukemia for physicians, nurses, and allied staff.
Implementation of the leukemia care program included multidisciplinary daily rounds, a weekly leukemia board, and 24/7 coverage in the inpatient unit by trained pediatricians.We recruited six pediatric oncologists and eight pediatricians over 5 years.Additionally, skill development and upgraded equipment for the pathology and hematology laboratories were emphasized as early priorities.
To build the PHOU's nursing workforce, we recruited nursing staff permanently assigned to the PHOU.The nurse educators taught a 6-week orientation course developed by St Jude Children's Research Hospital, 20

Block 3. Health Information Systems
We aimed to establish a system for the collection, analysis, application, and dissemination of reliable data to monitor clinical progress and effectively design data-driven quality improvement projects.To achieve these goals, we developed a hospital-based cancer registry at HGT and partnered with the Tijuana population-based cancer registry, BajaREG. 21

Block 4. Access to Essential Medicines
To ensure a consistent supply of equipment, medications, and supplies with priority placed on quality, safety, efficacy, and cost-effectiveness, we aimed to develop practices to reduce waste and garnered financial support from the HGT leadership and Patronato, a local grassroots nongovernmental foundation.

Block 5. Financing
To address financial sustainability of ALL management, we applied for national accreditation to ensure funding through the Popular Insurance Program (Seguro Popular) 22

Program Sustainability Assessment
Sustainability indicators encompassed three major domains: process, staff, and organization and were analyzed in 2013 and at the end of 2017 using a validated score-based sustainability tool. 23The National Health Service (NHS) Sustainability Model is a self-assessment tool detailing 10 key indicators that increase the likelihood of sustainability and continuous improvement for a specific change that has been introduced into an organization.Scores represented leaders' and point-of-care staff's perceptions.

WHO Framework for Action Implementation Results
The results of the implementation of the WHO Framework for Action model are summarized in Table 2.

ALL Clinical Outcomes
Approximately 20-22 children with ALL per year were expected in Tijuana on the basis of the childhood population.
The majority (approximately 60%) were diagnosed and received care at HGT.
In total, 109 children with newly diagnosed ALL were included in this study and were divided into two cohorts: preimplementation (n 5 49, from 2008 to 2012) and postimplementation (n 5 60, from 2013 to 2017).Patient characteristics are described in Table 3.

Sustainability Assessment Results
All the sustainability indicators assessed had an improvement potential when documented for process, staff, and organization domains (Fig 3).The two indicators of the process domain with the highest improvement potential were the adaptability of improved processes and the effectiveness of the system to monitor progress.Clinical leadership engagement was considered the most critical indicator in the staff domain.In the organization domain, the infrastructure for sustainability indicator was the most important (

DISCUSSION
The application of the WHO's health systems strengthening Framework for Action model and its six building blocks: service delivery, workforce, information systems, access to essential medicines, financing, and leadership/governance resulted in the successful implementation of an effective and sustainable leukemia care program in the US-Mexico border region.The 5-year OS and EFS improved significantly postprogram implementation.Moreover, we demonstrated improvement across all three NHS sustainability domains: process, staff, and organization.These results support that pediatric ALL outcomes in LMIC settings can be meaningfully improved through partnership programs designed according to health systems strengthening models.
3][4] Although causes of failure can vary across LMICs, the application of health systems strengthening models focused on building sustainability are essential to closing this survival gap.According to our sustainability assessment, improvement potential was achieved in 2013 after implementing the leukemia program at HGT and sustained into 2017 in 6 of the 10 sustainability indicators, enabling greater capitalization of opportunities to improve survival outcomes.These improvements in sustainability suggest that the improved clinical outcomes can be sustained long-term and allow the program to adjust and continue to make improvements to respond to new challenges in providing comprehensive leukemia care in resource-constrained settings.
A significant improvement in OS and EFS between preprogram and postprogram implementation for both patients with SR and HR ALL and a steadily increasing 5-year OS rate of 65% for the entire ALL cohort within a decade of our program's inception demonstrate sustained highquality care and improved outcomes.Additionally, an increase of more than 30% improvement potential across all 10 sustainability indicators from 2013 to 2017 suggests the success of the WHO Framework for Action approach.14 The scale of our model is small; nevertheless, there are precedents for endeavoring toward large-scale pediatric leukemia care improvements in LMICs.Examples include the Asociaci ón de Hemato-Oncolog ía Pedi átrica de Centro Am érica and Mexico's Seguro Popular initiative, which in 2008 introduced new coverage for pediatric cancer treatment.However, Mexico still faces challenges in providing effective ALL treatment.It was estimated that only 48% of eligible patients were covered by Seguro Popular, and clinical outcomes remain variable across different regions. 22For instance, studies report that 62% of pediatric oncologists in Mexico remain concentrated in its three largest cities (Mexico City, Guadalajara, and Monterrey). 22lthough two studies 11,30 report a higher prevalence (58%-78%) of HR ALL, similar to our rate of 67%, a lack Fig 3).In 2017, 4 of 10 sustainability indicators reached the maximum improvement potential, 2 of 10 sustainability indicators had improvement potentials below 10%, and the remaining 4 of 10 sustainability indicators had improvement potentials below 40% (Fig 3).
20stainable Pediatric Leukemia Care in Mexico by visits to Rady Children's Hospital San Diego and to St Jude Children's Research Hospital affiliated sites in Guatemala and Chile.20 including advanced pediatric leukemia clinical competencies.Training was supplemented FIG 1. Twinning-WHO framework combination model to achieve high-quality leukemia care.JCO Global Oncology ascopubs.org/journal/go| 3 Events included relapse, treatment abandonment (interruption of treatment of at least 4 weeks for nonmedical reasons), and death.The Institutional Review Boards for the University of California San Diego/Rady Children's Hospital San Diego and HGT approved this study.Informed consent was obtained from parents/legal guardians.
and established a partnership with Patronato.Additional startup funding was obtained from Rady Children's Hospital San Diego and St Jude Children's Research Hospital.immunophenotyping, genetic features, and measurable residual disease (Table 1).Two cohorts were analyzed separately: preprogram implementation from 2008 to 2012 and postimplementation from 2013 to 2017.Outcomes including OS and event-free survival (EFS) were estimated and stratified according to risk of relapse (standard-risk [SR] or high-risk [HR]) and compared between the two cohorts.

TABLE 2 .
WHO Framework for Action Implementation Results Block 6: Leadership/governance Assigned bilingual/bicultural pediatric hematologist/oncologist (P.A.) to oversee all efforts Completed memorandum of understanding between Rady Children's Hospital San Diego and HGT, developed a leadership team (operations and clinical staff) responsible for continuous improvement in clinical services, managing operating budgets, developed guidelines for consistent standards and staff performance expectations, and forecasting long-term plans Abbreviations: CLABSI, central line-associated blood stream infections; EVAT, Escala de Valoraci ón de Alerta Temprana; HGT, Hospital General-Tijuana; PEWS, Pediatric Early Warning Score; PHOU, Pediatric Hematology Oncology Unit. a PEWS or EVAT in Spanish.b Unpublished.c MEJOR (Spanish) means better in English.JCO Global Oncology ascopubs.org/journal/go| 5

TABLE 3 .
ALL Patient Characteristics From 2008 to 2017 Forthcoming initiatives should consider the implementation of standardized approaches for data collection through cancer registries and surveillance systems, the building of sustainability through financial and bidirectional leadership structures and the provision of patient-centered care to mitigate socioeconomic barriers to optimal clinical outcomes.Future research should evaluate best practices in establishing global health partnerships aimed at improving cancer care through the application of health systems strengthening models adapted to local health systems in each unique community.Staff involvement and training to sustain the process 2b.Staff behaviors toward sustaining the change 2c.Senior leadership engagement 2d.Clinical leadership engagement Organization domain NHS Program Sustainability Assessment Scores: 2013, 2017, and Maximums FIG 2. ALL survival outcomes.HR, high risk; OS, overall survival; SR, standard risk.JCO Global Oncology ascopubs.org/journal/go| 7