Golden Keys

According to the International Agency for Research on Cancer, more than 175,000 new cases of cancer in children (0-14 years) are diagnosed.   In low-income countries (LICs), cancer is the most common cause of death in the child population and cure rates are at least 20-30% lower than in high-income countries (HIC). In El Salvador, Central America, as in other low-income countries, socioeconomic factors such as poverty and minimal education are predictors of late diagnosis and advanced disease at diagnosis; particularly in solid tumors.  The overall 5-year survival rate for all solid tumors in HIC countries is around 80%, while in El Salvador it is well below 30%,  as most children are diagnosed with metastatic disease.

An increase in survival will be achieved when LIC countries identify the disease as early as possible, which is associated with less intensive treatment regimens, shorter treatment duration, fewer side effects, reduced treatment costs and fewer complications.

This has increasingly become a major public health issue in LIC countries in recent years,  particularly because it affects the pediatric population diagnosed with solid neoplasms because they are often diagnosed in advanced stages making surgical treatment eminently a challenge or  impossible.

Thus, with the implementation of policies for the prevention and control of cancer in its objective three proposes "To develop the institutional capacities of the National Health System for early diagnosis, timely treatment, rehabilitation and palliative care of oncologic pathology at different levels of the Redes Integrales e Integradas de Servicios de Salud (RIISS), for the provision of services in an equitable, integral, and sustainable manner."

 According to the Registro Oncopediátrico Salvadoreño (ROPESAL) the incidence of cancer between 0-19 years of age is 131 cases per million. Childhood cancer is the sixth leading cause of mortality in children under 15 years of age and the third leading cause of mortality in adolescents between 15-19 years of age.

In El Salvador, the majority of childhood cancer cases diagnosed (80%) are diagnosed late with advanced disease which contributes to treatment challenges and reduces the chances of survival. One strategy to reduce the diagnosis of pediatric cancer in late stages is to create a warning system in the primary health care system, which is the first contact of the patient in the community, which favors the identification of patients at risk and with red alert signs of cancer risk to promote an early diagnosis in early stages.

Pediatric cancer is considered  a rare disease and most primary care physicians or nurses rarely see a case during their career span. This lack of experience and short consultation time suggests that early diagnosis cannot solely rely on the care provider’s skills. Hence strategies that support the physician’s skills are needed. 

Our solution is to develop an addition to our already existing medical software(SAM), which provides safety and quality in chemotherapy administration, starting with the ordering of chemotherapy ; continuing with order verification and preparation in the pharmacy; and concluding with nurse administration.  It also allows us to make daily revisions of absentees to treatment as well as registers information from social work, psychologists, pharmacists, oncologists, and other disciplines.  The new addition will be a keylogger software for medical use installed in the primary health care system clinics that is capable of capturing key words associated with symptoms of childhood cancer and alerts physicians of a possible childhood cancer case.  The keylogger software will be linked to SAM. 

These phrases will be structured in a database of words to develop a semantic search that will be able to classify them and verify if they are related to childhood cancer. The symptom key words will be previously determined by pediatric oncologists and the multidisciplinary team.

Once these keywords are detected, the software will alert the physician in primary health care.  The software will have three main functionalities:

a) Provide contact Information: The software will provide telephone numbers and an e-mail of Fundación Ayúdame a Vivir, the only pediatric cancer program in El Salvador. The primary health care physician can contact the childhood cancer program so that they can be directly oriented on how to proceed with the patient (further analysis and exams to discard a cancer diagnosis). 

b) Registration Link: Where the primary health care physician can write in a form the patient's information (file number, patient's name, contact information, and a brief description of symptoms). This information will be directed to Fundación Ayúdame a Vivir, the organization that can capture the case, contact the patient, and promote referral to further evaluation. 

c) How-to-proceed notifications: The Software will recommend actions depending on the words or phraseology detected, which will contain cancer information duly structured and revised by medical oncologists that can guide the patient about their possible condition. For instance, if the software detects the phrase “abdominal mass” a notification will pop suggesting the primary care provider to order an ultrasound. 

This software will need little to none learning from clinicians at the first level. Its interface will be intuitive. Clinicians will not need to open additional windows during their daily consultations because suggestions will appear as regular notifications. Additional windows will pop in case they need to send a short form to notify Fundación Ayúdame a Vivir about the case or if they request additional management information. 

El Salvador is the smallest country in Central America, bordered by the North Pacific Ocean and located between Guatemala and Honduras; it has a land area of 20,721 km2, approximately the size of Massachussets. The Gross Domestic Product (GDP) growth in El Salvador reached 1.862 percent in 2019, but the country has registered persistently low economic growth levels due to violence and poverty. According to the International Monetary Fund, GDP per Capita is equal to $ 9,139.70 per year. The publication entitled "Projections and Estimates of the Population of El Salvador (1950-2050)", produced by the Ministry of Economic and General Office of Statistics and Censuses, reports the projection of El Salvador's total population for the year 2020 of               6,765,935 inhabitants. The pediatric population (0-18 years) constitutes 35%, comprising 2,387,091 inhabitants.

El Salvador´s National Childhood Cancer Program, supported by Ayúdame a Vivir Foundation in alliance with Children´s National Hospital Benjamin Bloom and St. Jude Children's Research Hospital, was established in 1993.  Since its beginnings, the program seeks access to comprehensive treatment for Salvadoran children with cancer, improving their quality of life and their family in all stages of the disease. The program has national coverage of pediatric cancer and receives 200 new childhood cancer cases annually and treats 350 patients per year in an age range of 0-<18 years. Treatment is free of charge for patients, but the disease incurs a lot of out-of-pocket expenses, like travel allowance. Leukemia (48%) is the most frequent pediatric cancer in our community, followed by Lymphomas (13%) and Central Nervous System Tumors (12%). The treatment for patients can last a minimum of nine months to a maximum of two years and a half.  According to socioeconomic interviews, 51% of our pediatric cancer population lives in extreme poverty and 49% of our patients live within a subsistence economy.  75% of children with cancer in El Salvador have to travel between 2-12 hours to reach the hospital, and only 25 % of the patients live in the surrounding cities near the hospital. 

Progress in childhood cancer treatment has resulted in survival rates of over 80% in High-income Countries (HIC). The reality is different in low and middle-income countries, where approximately 80% of children with cancer are diagnosed but only 20% are cured. In El Salvador, cure rates are 65%.  Abandonment of treatment rates in low and middle-income countries is high due to poverty, lack of access to treatment, and socioeconomic factors.  El Salvador has an abandonment of treatment rate of less than 1%, primarily fostered by education to the caregiver and a tracking protocol that allows healthcare personnel to detect absences in real-time. 

 Target population:

Our solution has the potential to impact every pediatric-aged individual that interacts with the first and second healthcare level across the country. Because every consultation is digitally recorded, every child that consults will be automatically screened for pediatric cancer. 

In 2019 (last data published by the Ministry of Health), 490,294 pediatric consultations were provided within the public network. Given that pediatric cancer emergence depends mainly on random genetic mutations, all children could potentially develop cancer and all children with symptoms should be screened properly. 

Our innovation  will have a direct impact on all children that consult in primary health care centers, especially those who will develop cancer. Because cancer outcomes are time-sensitive, early diagnosis will lead to better clinical outcomes in the long run.  As of today , children with cancer are not properly screened or referred to a tertiary health care center. These delays impact how their cancer is treated and children´s prognosis. 

One example is treatment intensity for Acute Lymphoblastic Leukemia (the most common pediatric cancer) . Treatment is dependent on several factors at the time of diagnosis, among them white blood cell count at diagnosis. A late diagnosis is usually associated with  higher white blood cell count and a more intense treatment is needed. Treatment delays lead to more intensive treatment which  have more side-effects, higher risk of infections, and deeper neurocognitive and functional impairment. Our solution will allow patients to arrive at our center in a timely manner and reduce the complexity of care and increase prognosis. 

Our innovation will accelerate the diagnosis  and give primary health care physicians the necessary information on referring a patient to the only pediatric cancer program in the country. This gives the opportunity to start chemotherapy, surgery, and/or radiation sooner. Detecting cancer in early stages will save lives and enhance quality of life of the patient and the family in a population that is exposed to chronic stress. 

Children with cancer are an underserved population.

The pediatric cancer program in El Salvador is funded mostly by Fundación Ayúdame a Vivir and it is an underfunded cause. In El Salvador, cancer care in general comprises less than 3% of the health care budget for all the population with most resources being allocated for adult cancer prevention and treatment. Historically, institutions tackling adult cancer have seen a growth in their infrastructure increasing the number of centers and services while the pediatric cancer program has increasingly improved its outcomes and dealing with increased prevalence with the same infrastructure and staff.  El Salvador´s childhood cancer program has been deemed an example of a cost-effective program by the World Health Organization. 

The team has been very involved in searching for solutions to deliver the best quality care to pediatric cancer patients.  The team is composed of a multidisciplinary group that respects each other professionally and is open to each person's contributions. 

Our team has experience designing technological solutions for pediatric cancer care. The software used at our program has been conceived, designed, and implemented by our team. This software includes modules for every discipline. A highlight of this software is that it has helped the team ensure chemotherapy safety, by ensuring it is administered to the right patient at the right dosage. Our system keeps growing every year, adding more features and functionalities. Another important contribution is that the software developed by our team has been a valuable tool to track absences  and helped reduce treatment abandonment from 13% to around 1% (Salaverría et al, 2015). Abandonment of treatment is the most severe form of non-adherence and a negative predictor of clinical outcomes. 

The particulars behind each idea are constructed by the oncologists, psychologists, the IT Team, and a nurse educator. Each person brings in a different view of the problem with the necessities each may face from their own line of work.  The execution of the software creation is the responsibility of the IT software engineer with the aid of the nurse educator, who has been extensively involved in the creation, modification, and execution of the program´s software.  Many successful additions of the software have been done this way, with each person contributing with ideas and counseling. 

The oncologist, psychologists, and nurse educator have contact with the pediatric families every day and we have heard each patient's story of how they came referred to our center. We know by their storytelling the delays in diagnosis, the weeks or months they spent consulting with different physicians, and the out of pocket expenses to pay for private physicians and exams. Most of all, we know of the guilt and responsibility parents feel when they are finally diagnosed with cancer and how they ask themselves if they should have done more. 

Childhood cancer is a rare disease and has become one of the leading causes of death in children. There is a global inequity in its treatment. with 20% of childhood cancer cases being diagnosed in high-income countries, while 80%  come from low and middle-income countries where access to treatment is challenging due to financial difficulties, distance to treatment centers, countries lacking the capacity to provide quality care, unavailability of medicines, among others. El Salvador is a low-middle income country with success in its program. The oncology team, from the doorman to the oncologist, is devoted to the project.  We have implemented projects that benefit our patient population with little resources, being an example globally of a cost-effective childhood cancer program.  We believe we can always do something more to better our service and attention to the pediatric cancer population.

We strongly believe that saving children with cancer is a nation-wide responsibility. Primary care  attention in health plays a key role in detecting and referring patients in early stages of the disease. But the reality of the systems lays a large burden or responsibility on the physicians who experience work overload. Hence, systems that facilitate their work and act as second layer of criterion need to take place in the first and second level. 

This challenge offers an opportunity to apply a technology solution to improve survival rates of a  rare childhood disease.

Our solution aims to accelerate early detection of cancer among children in El Salvador. It is innovative because the solution targets all children in the country who consult in primary care centers, it uses technology to facilitate detection of symptoms, and creates a connection between primary care physicians and specialists to improve quality of care.  This innovation will be implemented at the same time as two other interventions to address early detection of pediatric cancer symptoms. Early detection of cancer is primarily targeted through education, which will be done through a project that intends to educate physicians through cascade teaching and a national public awareness campaign of childhood cancer symptoms.

By allowing primary care physicians to identify symptoms through the keywords, we are empowering them to respond promptly to signs of possible cancer cases. We are offering the pathway  on how to proceed to do further testing and make a faster diagnosis of cancer. 

During the first year, our goal is to develop the software, make an alliance with the public health system, and incorporate the solution into primary health care centers through education. During this period we expect to have executed a pilot test and initial referrals to the program arriving. The IT team is already working on the concept but the funding is needed to begin its development. Alliances with the public health system have begun, in the development of the National Cancer Control Program. 

 During the next five years, we envision Golden Keys functioning normally within the public health system. We believe with this innovation in work,  referrals for pediatric cancer will be made sooner and patients will lose less time. In five years, diagnosis delays should decrease and survival rates for pediatric cancer should increase. The program should also have a decrease in financial burden, having invested less in chemotherapy, support medicine, and interventions in complicated patients. 

• Number of children screened for suspicion of childhood cancer through Golden Keys.

• Time it takes for a patient to consult with the primary health provider, measured in days.

• Time it takes to be referred to tertiary hospital since the first consultation, measured in days. 

• Number of consultations made by primary health providers to the oncology team through the software. 

• Number of cancer diagnosed after prompt intervention using the software.

• Number of Lymphoblastic Leukemia patients diagnosed with less than 20,000 white blood cell count, making them candidate for the low risk treatment arm. 

• Number of patients diagnosed with solid neoplasms referred through Golden Keys that at diagnosis surgical frontline management is an option. 

• An improvement in knowledge about childhood cancer evidenced by an increase of consultation to the oncology team or referrals through the secure form.

In El Salvador, 80% of pediatric cancer patients have a late cancer diagnosis. This rate was determined from an analysis made by -(ask Soad) and is supported by data from interviews with caregivers of cancer patients. There is a low efficiency to detect cancer and low recognition of signs and symptoms of pediatric cancer among primary healthcare providers. Our objectives are to increase awareness of childhood cancer, increase efficacy in detecting signs and symptoms of cancer, and make early detection and diagnosis of childhood cancer. 

Our inputs will be the following:

1. Funds: Funding will be necessary to achieve our goal. Investment will be focused on paying developers and supervisors fees, contracting cloud services (hosting, database, ssl certificate and backup), buying technical equipment such as programming computers, sponsoring a dissemination campaign, and producing an instructional video. 

2. Technological resources: 

   1. Existing resources: Golden Keys will act as an extension of our current system (SAM). 

   2. Prospective resources: Include databases subscription, server purchasing, programming hardware and software, etc. 

3. Collaborative partnership: Include internal and external collaboration as follows:

   1. Stakeholders support: We require approval from executive leadership and IT staff at the Ministry of Health to install Golden Keys on the computers at the clinics in primary level of care. 

   2. Communicational coordination: Communication between PR departments at Fundación Ayúdame a Vivir and the Ministry of Health to promote the installation and use of Golden Keys at the primary level clinics. 

4. Human resources:

   1. IT Staff: For this project hiring an additional developer is required. Current IT staff will assume the responsibility of supervising the development of Golden Keys and representing Fundación Ayúdame a Vivir during coordination with IT staff at the Ministry of Health and its allied clinics. 

   2. Clinical staff: Composed of medical, nursing, and psychologists  will be responsible for reviewing patient’s files and creating the bag of words that the system will use. This staff will also be responsible for interviewing families who are referred through Golden Keys. 

   3. Communications staff: Will assume the responsibility of creating a dissemination campaign and producing a video to inform first level clinicians about the existence and use of Golden Keys. 

5. Knowledge: SAM was developed with several inputs and our patients being the center of the customizations. Knowledge from clinical and IT areas will be key to deliver a well-informed, amicable,  and valuable software. 

6. Patient’s files: Every child with cancer has during their disease journey loads of data that is safely guarded physically and digitally by our team. This data is the most important source of words, abbreviations, and phrases for the bag of words. 

We will use these inputs to develop the following activities:

1. Meet with stakeholders to create a strategic alliance: We will welcome executive leadership, IT staff,  primary care physicians, nurses. and health promoters at our center to discuss with them the conditions in which Golden Keys can be implemented. Their opinions and expertise will be considered to develop a secure and responsive software. These meetings will take the format of a workshop based on design thinking methodologies that will allow both parties’ needs to be represented when building the software. 

2. Interview families: We will speak further with families to better understand their trajectory until final diagnosis of pediatric cancer.

3. Review patient’s files: Patient’s files will be reviewed by the clinical staff. The psychology team will be responsible for identifying commonly used words to describe the medical semiology related to pediatric cancer in files of patients that came to our program: (1) critically ill, (2) in advanced stages, (3) patients who came in early stages, (4) patients who came in “average” time. For this purpose, referral notes (notes written by physicians at first and second level clinics or other services to refer the case to our program for further study), admission notes, evolution notes, radiology and laboratory reports, nursing notes, and other initial case annotations that can provide information as what was the semiology used by primary care physicians. This information will be presented to the oncology team, who will be responsible for writing the recommendations that will appear as notification when certain words are typed, for example, if the word “blasts” is typed, the system will recommend a peripheral blood smear.  Oncologists will also present sets of signs and symptoms that found together are a typical indicator of certain types of cancer that require immediate response. For example, if “leukocoria” and  “glaucoma” appear together (typical signs of retinoblastoma) an alert will appear suggesting that this particular case needs to be referred to an ophthalmologist. This recollection of words, phrases, abbreviations that represent typical signs and symptoms of pediatric cancer will be the contents of the bag words that the developing team will receive and load into the database.   

4. Write the source code: Our IT team will be responsible for coding Golden Words with the information (words, abbreviation, and phrases) collected during the patient's files review. After coding, they will execute a pilot test to ensure all functionalities are properly working. 

5. Disseminate the new software: With the Golden Keys v.1 software available, a coordinated effort between the Ministry of Health and Fundación Ayúdame a Vivir will install the software on a initial number (to be determined) of clinics to gather more data and include improvements before nation-wide dissemination starts. 

6. Produce a presentation video: To present Golden Keys to clinicians and provide information about how to interact with the software in case referral is needed. FAV´s communication department will be responsible for this activity. 

Having executed these activities, we expect the following outputs in 6 months after starting:

1. Bag of words: Collection of medical semiologic expressions about pediatric cancer signs and symptoms.  

2. Golden Keys v.1 (Keylogger software): software that traces the computer’s keystrokes in the digital patient’s charts. 

3. Secure form: A web-based form to refer patients providing relevant clinical data and contact information. 

4. SAM’s Golden Keys module: An internal module at Fundación Ayúdame a Vivir software to receive clinical and contact data generated at the secure form. 

5. Promotional videos. 

Based on these activities  we expect the following outcomes:

1. Short-term outcomes (6 months after Golden Keys launch):

   1. At least 25 primary level clinics have installed Golden Keys in their computers.

   2. A recall  of the importance of early detection among primary healthcare providers.

   3. Increase in awareness of childhood cancer among primary healthcare providers. 

   4. Consultations to the pediatric cancer program through golden keys should begin.

   5. Increasing knowledge about the software's existence between primary care providers. 

2. Intermediate outcomes (2-3 years)

   1. At least 50% of all specialized health units nation-wide are using Golden Keys.

   2. At least 50% of referrals to the pediatric cancer program are made through Golden Keys.

   3. A sustained increase of patients with solid neoplasms referred to the pediatric cancer program through Golden Keys arrive with better prognosis. 

1. Long-term outcomes (4-5 years)

   1. At least 75% of specialized and 50% of intermediate health units have installed Golden Keys. 

   2. At least 65% of referrals to the pediatric cancer program are made through Golden Keys.

   3. An increase of 50% of patients referred through Golden Keys come to the pediatric cancer program with no metastatic disease. 

Our impact goals are as follows:

1. Increase of 5 year event-free survival for patients who were referred using “Golden Keys“ compared to patients that were referred through other mechanisms.

2. Increase of cost-effectiveness of the pediatric cancer program in El Salvador. 

3. Reduced time of diagnosis since onset of symptoms in the overall pediatric cancer population. 

The search engine for words or phrases related to childhood cancer will be based on natural language processing using the "bag of words" method as the backbone of the application. The bag of words is a representation of a text that describes the occurrence and frequency of known words. In this case, all words related to cancer signs and symptoms. For this we will store in a  volatile memory the phrases that will be captured by implementing a keylogger, where we will assign each word a numerical value comparing it against a vocabulary of genuine words related to childhood cancer symptoms. This will generate a set of probabilities that through the "Bayes" method allows the classification of  the probability of cancer. At the end of the procedure, it is coupled to a cloud controlled and reviewed by the cancer professionals in El Salvador of the Organization Help Me Live Foundation. The oncology team will give the respective follow-up for each alert.

NLP: Natural language processing (NLP) is a field within artificial intelligence and applied linguistics that studies interactions between humans and machines using natural language (we will use the field summary and text classification).

Keylogger : Tracks and records every keystroke on a computer, often without the user's permission or knowledge.

Bag of words : The bag of words model is used to preprocess text into a bag of words or vectors of fixed length, a machine learning algorithm. It is the simplest form of text representation in numbers.

Bayes' theorem:  Is used to calculate the probability of an event, having information about that event beforehand. We can calculate the probability of an event A, knowing also that A fulfills a certain characteristic that conditions its probability. 

The project team has worked together for many years and on many projects, sharing common goals, expectations, and a passion for the pediatric oncology program. We have respect and appreciation for the contributions each team member delivers and with this respect for each other comes a fair distribution of power, recognition, and risks. We share the same values and goals: union and passion, respect and compassion, and honesty. We belong to the same team and we work together to achieve our mission to help all children with cancer in El Salvador live. Each one of us is a key piece to achieve our goal and we are united by our passion for the cause. We believe we are equal. That is why we treat all of our patients, staff, volunteers, visitors, donors, and stakeholders with respect, kindness, and compassion. We are committed to the responsible and ethical management of our resources. 

The pediatric cancer program in El Salvador was initiated and is presided by Fundacion Ayúdame a Vivir(FAV), a non-profit organization that provides treatment and care for all Salvadoran children with cancer, who would otherwise not be able to pay for the treatment. The impact of the program is saving the lives of hundreds of children each year. With St Jude Children's Research Hospital and the Ministry of Health as stakeholders,  FAV can offer a comprehensive cancer diagnosis within days, all chemotherapy free of charge to patients, psychosocial support, and lodging for patients who come from outside the city, among others.  Human resources have been trained in pediatric oncology, including oncologists, nurses, and psychologists to deliver a comprehensive standard of care to all Salvadoran children with cancer. The majority of patients are unable to pay for cancer treatment and there is no treatment in private hospitals in El Salvador so their child's chance of survival is made possible by FAV´s work. Funding for the national cancer program comes from the government, through the Ministry of Health, and fundraising by FAV. From this funding comes all chemotherapy, antibiotics, and staff salaries. Patients from all over the country are referred by primary and secondary level hospitals and once a cancer diagnosis has been established, they are immediately admitted to the program. Patients might have out-of-pocket expenses, like travel expenses, and this is also sometimes covered by FAV and its allies. 

Financial sustainability can be achieved in two different stages. In the first stage, funding comes from grants and is oriented to building, promoting, and implementing our innovation. This funding is key for developing a strong, secure, and amicable program. Testing is also important to make adjustments. 

When the software is up and running and results as well as indicators are achieved, we plan a two-way strategy: this software can be sold as individual licenses to other pediatric cancer programs in the region and worldwide as well as to governments who are willing to increase their capacities to detect pediatric cancer in early stages. 

The other strategy is working collaboratively with other medical specialties that could benefit from early diagnosis. This is to add additional modules to detect other time-sensitive diseases. Additional modules can be sold separately as plug-ins to institutions such as Fundación Ayúdame a Vivir or governments interested in supporting their clinical workforce through technology.  

This aligns with the WHO's Global Initiative for Childhood Cancer (GICC), whose target is to achieve at least 60% survival for children with cancer by 2030 by involving governments and increasing countries’ capacity to provide quality services for children. Different levels of subscriptions to the software can be offered to fit the needs of any program.

Fundacion Ayudame a Vivir is a 30-year program that has received funding from different sources. Some examples include funding from Fundacion Gloria Kriete, which provided funds for a Flow Cytometry Laboratory valued at approximately $100,000, which provides a timely diagnosis of pediatric patients with leukemia. The leukemia immunophenotype determination allows the treatment assignment based on the risk stratification of the disease. Before acquiring the Flow Cytometry, diagnoses were made in Guatemala and took up to 3 days.  Today we can make a leukemia diagnosis in 24 hours. 

 In 2013, grants totaling more than $636,000 over two years were awarded to two physicians. One of them was Dr. Soad Fuentes Alabi, who studied the epidemiology and biology of childhood cancer in South America and worked to establish a cancer registry in the region. She set up a cancer registry in Central America, demonstrating that the incidence of childhood cancer was higher than previously thought and showing that, as in other countries, governments need to allocate more resources to cancer programs. 

 Another important grant of $100,000 was given in 2021 for infrastructure. With this funding, the day patient clinic was completely renovated. Built in 2008, the day clinic had become small for the patients and personnel. Caregiver's bathroom areas were improved, a resting area was created for caregivers, an area for adolescent patients was made, a school space was renovated, office spaces were created for staff who had no working areas, and the play area for children was improved.