Sehat Sanjha

Problem definition: 

We will be addressing a two-fold problem centered around information asymmetry in healthcare. The two prongs of this asymmetry are: 

• Doctor-patient power imbalance: Hospitals and doctors’ offices have more knowledge about the patient’s condition than the patient themselves. 

• Navigating the industry as a patient: The navigation path through the healthcare system for treatment of a serious disease is difficult, more so in developing countries.  

Problem description:

First prong: 

Lahore is a city of 11.13 million people and with more than two dozen tertiary care centers. Good data for the numbers of smaller hospitals, general practitioners, specialists, dentists, pharmacies, and procedure centers is not available but there is a broad consensus that Lahore is not underserved in terms of the availability of health resources. According to a Lancet study, Pakistan ranks 154th among 195 countries in terms of quality and accessibility of healthcare. The problem is not that the resources do not exist, it is accessing these sources in a timely manner that proves to be a hurdle in optimal patient care.

Not just in Lahore, but globally, information asymmetry is a core problem in the delivery of healthcare. Hospitals and doctors’ offices have more knowledge about the patient’s condition than the patient themselves. And it is impossible to weigh the relative weight of a piece of medical information unless you are a trained medical professional. This creates a power dynamic in doctor-patient relationships which is not always congruent with optimal outcomes for the patient. Patients do not know about the competence, fidelity in delivery, bedside manners, and prices of healthcare providers. This creates a gap that leads to lower patient satisfaction and quality of care.

Second prong: 

Most Pakistani healthcare facilities do not have arrangements to maintain patient records. The patients themselves may bring a stack of documents with them to appointments, but that is a rare occurrence. Thus, virtually all care is event-triggered and/or emergency-driven.

When you engage with the healthcare system only when you or your loved one is already sick, and the system has no social workers or care coordinators to guide you, navigating the system is hard which means many patients hesitate to initiate care, downstream care decisions are delayed and as a result, patients have suboptimal outcomes and increased morbidity and mortality.

This leads to two main problems:

1.  History-taking is often summarized and rushed

2. This makes navigation through the system difficult.

Lahore has government hospitals where care is technically free, and although these hospitals are overcrowded and often have shortages of medical supplies, our research has shown that the number one complaint patients have is that the hospitals have no organization to guide them on where to go for their problem. Moreover, calling a hospital or a doctor’s office or an emergency room in Lahore is notoriously unhelpful. Many places do not have phone lines that are manned round the clock.

Value relative to existing solutions: 

In 2021, the Government of Pakistan started a health card that allows each Pakistani to get PKR 400,000 ($2200) worth of care in the event of an illness serious enough to need hospitalization. However, there is still no searchable directory with information on a doctor’s specialty, their experience, their locations and hours of service, their ranking as per previous patients, or the outcomes of their caregiving.

Companies like Marham and Ola Doc do have helplines and websites with search functions where you can find a private physician in the specialty you are seeking. However, our research (patient interviews) has revealed that physician reviews on these websites are not very helpful. They also do not have information about government physicians, hospitals, emergency rooms, or procedure centers. Moreover, the models used by these services translate the intrinsic chaos from an analog setting to a digital world.

We are building a registry of medical resources in Greater Lahore, that will give patients up-to-date information on the current infrastructure to bridge the power imbalance between patients and doctors and help them navigate the system more effectively. 

We are solving the problem of information asymmetry and difficulty in navigating the healthcare system by building a registry of all available healthcare providers and facilities in Lahore. Imagine Google maps with Amazon product descriptions and reviews from previous users but for healthcare. Pakistan has 100M+ smartphone users and the numbers are doubling every other year. Our registry will sit on patients’ phones in the form of an app for patients to look up healthcare facilities and providers, with details about the provider’s specialty, timings, qualification, competency, location, and ranking as per reviews by other patients. This will help patients choose the providers most appropriate for them and hold the healthcare providers accountable for the services they offer.  

Our healthcare registry will help its users find the care they need by the following matrices. 

1. Where - Where is the doctor/hospital located in relation to where the patient is  

2. When – When are the hours of operations and when are phone lines open to access a particular doctor or procedure center  

3. Specific expertise + competency – What are the credentials of the doctor/hospital for providing care for my particular problem  

4. Ranking hospitals and doctors per reviews from patients who have received care from them  

Sources for the Registry 

1. Sehat Sahulat Administration  

The government of Pakistan launched national health insurance in 2021 called Sehat Sahulat. Sehat Sahulat has a paper registry of inpatient facilities in Lahore which is not complete but they have verifying physicians who go to hospitals and verify the institution and physicians affiliated with it. We are talking to Sehat Sahulat administration to get access to their registry so we can crosscheck it against our other sources and make our registry more complete. Sehat Sahulat does not have information about outpatient care because it only covers inpatient illness.  

2. Pakistan Medical Commission  

Pakistan used to have Pakistan Medical and Dental Commission which was dissolved in 2019 because of its poor functioning. Pakistan Medical Commission was started at the same time and although it gives web verifications for physicians who are registered with it, many physicians are not registered because transfer of data from PMDC was not robust and registering with PMC is so hard that you have to go to the office in Islamabad at least 2 times when you get registered. They do not believe in answering phone calls or emails. We will use PMC to verify qualifications for those physicians who are registered with it. 

3. Pharmaceutical Industry  

From interviews with physicians, pharma executives and hospital administrators, we know that the pharmaceutical industry in Lahore has the most complete lists of outpatient offices because their drug representatives use these lists to market their medications to the maximum number of physicians possible. We are talking to Pharma companies to see if we can access their lists.  

4. Google search  

We have started by designing search tools that scour the internet for names, addresses, hours and services offered at healthcare facilities with a presence on the web. Less than half of all healthcare facilities, especially if they are on the smaller side, have a website and the websites are often out of date.  

Many more facilities have Facebook pages so we will scour Facebook also. We will crosscheck FB against website against Yelp presence against Instagram accounts both to verify the facilities but also gather as much information as we can via web searches. 

5. Street by street survey of some areas 

The gold standard for any registry is street by street survey of an area which we will use to spot check the information we are gathering by other means. We will aim for >99% congruence between our registry and the real world and discrepancies greater than that would trigger a re-verification of the area. 

We have chosen not to address pricing at this point even though it is a big part of the black box nature of getting healthcare. We are not addressing pricing at this stage because asking for a pricing structure will make healthcare vendors hesitant about sharing any information with us and we want to start by building a robust digital model of the healthcare infrastructure in Lahore.  

We will prompt patients when they are writing reviews to tell us about the affordability of care. Very few people in Lahore have private insurance so most payments are out of pocket and physicians and hospitals often offer varying price menus to patients based on their ability to pay. We will ask patients to share prices in their reviews but we will not offer “official” pricing in the first version of the app. 

 6.Hours of service: 

We will call each facility and office on our registry to confirm hours of operation we have picked up from the web.  

Patients will be asked to confirm hours of operation when they are asked to review the physician/facility they have seen.  

7. Physician/Facility Qualifications: 

On the back end, we will have a record of what the facility/doctor says their qualification and experience is.  

Physicians who are registered with PMC will be verified via PMC. 

For other physicians, we will ask for references of people who went to medical school with them and try to verify their degrees via peer crosschecking.  

For facilities, the Health Commission in Lahore has a list of inpatient facilities which is not complete and has no information about the quality of the hospitals. We are talking to them but will depend mostly on facility visits for our due diligence.  

8. Reviews 

We know getting accurate and candid reviews is critical in creating a registry that is truly useful. We are developing a system in which patients can speak their reviews in English, Urdu and Punjabi and the registry will transcribe them so we can use them to effectively rate facilities and physicians.  

We will keep a record of all the searches a patient is doing and the quality of their reviews so we can give more weight to a ‘good’ reviewer and less to someone who maybe in an outlier in their experience. As we gather more data, we will start using machine learning models to convert patient reviews into quantified provider rankings which are accurate and fair.  

9. History of Individual User Behavior  

We will maintain a history of each user to see their search patterns and health needs. The search history, the timing of reviews and the quality of reviews will help us figure out which user has good information to share, and we will give their reviews more weightage in ranking a facility/Physician. 

Approximately 72% of Pakistan’s population owns or has access to a smartphone and we can estimate a higher figure for Lahore as it is the second largest urban hub of the country. Our first version uses English as the primary form of communication but we will add Urdu/limited literacy versions in the future as well. Since our solution is app-based, it will serve anyone who seeks healthcare given that they own a smartphone and are familiar with the English language. We do not believe internet availability and affordability to be a major hindrance for most smartphone users in Lahore. 

Our target demographic faces two main barriers with respect to healthcare - they are either unable to afford suitable care or they do not know where to go when faced with illnesses or emergencies. There is also information asymmetry, where patients are forced to take the healthcare provider's word at face value, leading to a power imbalance and lower healthcare satisfaction.  

The first barrier is catered to by the Government of Pakistan's Sehat Sahulat program where the availability of insurance eases the financial burden of healthcare. However, the second barrier is largely left unaddressed, and our solution aims to change exactly that. Our registry and review system will essentially close the loop for efficient and satisfactory healthcare by connecting resources and people.  

Our review system plays an especially pivotal role in the way our solution will impact our target populations' lives. By allowing both oral and written reviews on the basis of their comfort, we will ensure that the patients leave accurate and timely reviews of their experiences. In order to holistically incorporate patient feedback, we will rate their reviews on the basis of quality and identify outliers through data analysis. Our users, who have reported a lack of information access, will now be able to make informed decisions regarding their medical appointments and visits. 

Our team, albeit small, is well-balanced as it includes both technical and medical staff, as specified below. 

The team lead, Sheza Munir, is a junior undergraduate student at Lahore University of Management Sciences (LUMS) majoring in Computer Science, minoring in Psychology and is currently on the Dean’s Honor List. She has experience working on both research and programming projects related to Human-Computer Interaction (HCI) and Machine Learning, including tasks such as conducting user research, designing applications, and writing code in languages such as Python, JavaScript, HTML, and CSS.  

The second team member, Imama Zahoor, is a junior undergraduate student with a major in Computer Science and a minor in Psychology at the same university.  She is also on the Dean’s Honor List. She is proficient in multiple coding languages and has experience with designing and building applications and websites. Having taken Artificial Intelligence and Human-Computer Interaction as university courses, she is also knowledgeable about common machine learning algorithms and designing systems in a user-centered manner. 

Dr. Basmaa Ali is the third team member. She went to medical school at King Edward Medical College in Pakistan followed by a residency in Internal Medicine at the University of Illinois at Chicago. She completed her Masters in Business Management from MIT Sloan School of Management and is currently serving as an Instructor of Internal Medicine at Harvard Medical School. She has founded multiple companies, the last one being Zanjabee Integrative Medicine, and built MigraineAid, a data registry for predicting migraine with a colleague at MIT.  

An amalgamation of these skill sets has allowed us to come up with an innovative solution and positions us well to implement our plan. We have access to physicians and hospital personnel, to help us gain insights into the workings of major health care facilities and providers in Lahore.   

Our team is well-suited to deliver this solution because we speak the local languages of Lahore, Urdu and Punjabi, as well as English, which has helped us communicate better with the hospital staff and the target population of this solution.  Since our team members have lived in Lahore for many years, we are well aware of Lahore-based healthcare services, their operations, and shortcomings. Therefore, we can look at the solution from the vantage of a consumer and understand the context in which it will be operating in.  

  We are applying to Solve because of the following reasons: 

1. Lack of digital healthcare registries is a very common problem in many developing countries. Navigating healthcare is hard even in sophisticated economies like the US and western Europe. No one has built a registry the way we are building one but people in Mexico, India, Bangladesh, and China are trying to solve this problem in their own way. We want to use Solve as a bridge to connect with mentors and other teams from similar countries.  

2. Other industries have built sophisticated registries with millions of products and developed a common language that most of us understand. Amazon e.g. has a digital language as to where the cart is, how and where to read reviews, vote reviews up and down, product specifications, user photos, and so on. This is a complex environment that Amazon has trained us to use effectively. Facebook and Instagram, although far less complicated, have done the same. We want to build the Amazon of Healthcare for Lahore and we know we need to learn from other industries. Solve will help us connect to the right people in other industries.  

3. Machine learning applications in healthcare data have become a specialized niche. We need expert support in this area and hope that Solve can connect us to the relevant specialists.  

4. Building this registry is going to be expensive. We know Solve cannot fund this registry completely but being selected will give us legitimacy which will allow us to raise funds from other sources.  

Sehat Sanjha combines an element that users are familiar with (maps) with another element that they need information about (details about providers). Using a mental model that is already ingrained in the user’s mind will provide a seamless transition to understanding the new service. Instead of translating the chaos of the analog system into its digital equivalent, we are reimagining the whole workflow and making it intuitive, accessible and easier to navigate.  

The registry will hold providers accountable for their services as the users will be able to see previous patients’ ratings about the doctors’ competency, bedside manners, and (if reported) outcomes from the patients’ perspective. Providing this transparency allows the patients to analyze the quality themselves and make informed decisions. Existing solutions are limited to doctors’ offices, but Sehat Sanjha provides information about all aspects of the healthcare infrastructure, including emergency care centers, private and government clinics, specialty care centers, labs and imaging facilities, etc. The registry will create a digital map of the existing infrastructure and virtually transform it into one large Lahore-wide digital hospital. Connecting analog and digital systems is the first step in improving resource allocation. 

The impact brought by the registry will change the healthcare industry for the better. When there is an accountability source, a place where patients can see for themselves which provider is better for them, it will encourage and incentivize the providers to improve and keep up. Reporting systems for physician conduct are almost non-existent in Pakistan, with new committees created and dissolved every few years. We are using social power to rebuild the system, which is innovative in itself as it was not possible a few years ago. 

Second, it will ease the navigation process by distilling the complexity in the current system and guiding them to the exact place they need to go.  

This registry is the first building block in our plans for Sehat Sanjha. It builds upon current working infrastructure and will itself be used to build a hub of connection between patients and doctors. Future additions include electronic medical records (accessible by patients and doctors), a scheduling app to make timely care easier, data analysis by machine learning models for tracking patient outcomes. Our research has shown that while having a registry is great, health coordinators and attendants are extremely impactful as well. We plan on training these individuals to strategically improve patient care by accompanying patients to their appointments, especially when an attendant from the family is not available.  

Our immediate goal for the upcoming year consists of building a robust registry for Lahore that contains information regarding hospitals and providers and will be continually referenced and updated. This requires data streams from multiple sources such as healthcare providers directly and cross-checking and consolidating information from existing regulatory authorities (Sehat Sahulat) and ancillary partners such as pharmaceutical companies.     

The next step is to design and launch our application and gain 20,000 active users in the first year.  Coupling this with a review system (integrated within the application) will allow users to rate their experiences for future patients to make informed decisions. Further enhancement includes linking a patient’s previous searches at the back-end system to ensure genuine reviews and more efficient appointments.  

In the second year and onwards, the natural progression of the project based on our user research will allow us to incorporate an off-the-shelf EMR and scheduling software to facilitate medical appointments within the healthcare facilities, including medical and analog triage algorithms, data, and ancillary analysis and in-built templates for vendors. Over the next 4 years, we will engage over 100,000 active users and establish diagnostic related groups (DRG), preventive care, as well as API with main labs. The maturation of intelligent algorithms will ease expansion to other areas outside Lahore and include more languages, such as Urdu. 

We know that we have to make a sticky app for the registry to have an impact on the lives of people in Lahore.  

Not only does the registry have to be complete and accurate – no mean feat – but it has to communicate the information it contains in an intuitive and seamless manner. 

Our matrices for measuring our progress towards our impact goal are as follows: 

1. Registry completeness 

Complete our registry by the end of year 1. 

Spot check the registry against street-to-street survey and our aim is for 99% congruence with the real world in one year for the city of Lahore.    

Compare our patient reviews with industry interviews of the doctors and facility managers to get a sense of the accuracy of our reviews. At least 5% reviews every year.  

2. App usage which will reflect the success of app design  

We want to serve 15,000 people by the end of year one.   

We want at least 20% of patients to open the app at least one more time after downloading it and spend at least two minutes on the app.     

We want 10% of our users to do more than one search by end of year one. 

We want 10% of our users to leave a review if they have seen a provider.  

3. Matching resources to needs 

Over time we will get a sense of what people in Lahore need by their search terms and what is available from our registry. We will do our first matching of resources vs needs at the end of our second year.  

1. App: Our solution will take the form of an application through software development.  

2. Internet: The accessibility and scalability standards will be maintained by linking it to the internet – patients all over our catchment area will be able to use the app to make informed decisions regarding their medical appointments and visits.  

3. Cloud Database: Similarly, our cloud-based data hosting will ensure flexibility. 

4. ML Models: Suggestions for relevant healthcare providers will be supported by Machine Learning models where data about the patient and their input fields will lead to suggesting healthcare providers that are relevant to them. In the future, as we add more features to our service, we will develop new and improved algorithms to facilitate improved user experiences.  

5. Voice recognition and speech to text: Transcribing the patients' spoken reviews.

Our core team is made up of 3 women. Women leading an endeavor that sits on the cusp of technology and healthcare is ground breaking in most of the world. We live in a Muslim majority country where only 13% of girls make it to 9th grade.

We are aware of the privileges that allow us to do what would be harder for women and men who have not had our privileges. We all come from Punjab which is the most influential province in Pakistan and we have all had the advantage of coming from a middle-class home which valued education and have worked our ways to the best institutions in Pakistan and the US.

Our work culture philosophy is that we want talent and dedication to our mission. We want better outcomes for everyone,for our patients, hospitals and healthcare providers and our people at Sehat Sanjha. As long as a person  is diligent towards improving the lives of patients in Pakistan, we will make sure that their religion, gender, socio-economic status, physical ability status, child-care responsibilities, stage-in-life-committments ( pregnancy, older parents who needs care) are not hindrances in having them work with us.

We will make a special effort to recruit Christian, Sikhs and Hindus who make up ~ 2% of Pakistan's population. We will also go to Quetta, Peshawar and Hyderabad and Thatta to recruit from provinces that are not Punjab in Pakistan. Once we are a cash-flow positive company, we will support scholarships for women in provinces other than Punjab at the college level, so we are enabling the ecosystem that will produce future leaders for our organization from all over Pakistan and from both genders.

We strongly believe the registry should be free to use for patients. And we cannot take money from healthcare vendors for making the registry because that will tarnish the credibility of the information we are putting out.  

Building the registry is not cheap but it is not expensive also. We need $100K to build the registry and set up a scalable back-end software architecture for its expansion and updates. We will need money to attract the first 50,000 active users as well as fine-tune our review gathering and verification workflows via voice recognition and machine learning models.  Thereafter maintenance of the registry is not very expensive.  

We plan to raise the money for building the registry and acquiring our first 50,000 users via grants and gifts. Cost of getting a new app user in Pakistan is between $5 to $7. We will need a marketing budget of $300K in year 2 when we go all out to increase our user base.   

Thereafter, we will raise the money we need for operations by allowing intra-app advertisements. Pharmaceutical companies are not allowed to advertise direct to the consumer in Pakistan but healthcare facilities and doctor’s offices can advertise their services. A large market of ancillary health products – health insurance, government programs, durable medical equipment and home health services – will all be interested in advertising within an app that has 100K+ users looking for health services. These advertisement will be clearly marked as paid promotions and users will be trained by consistent placement to distinguish registry rankings from paid promotion.

Once we introduce Electronic Medical Records (EMR) and scheduling services in year 2, we will ask the vendors who are getting patients via out portal to give our patients a 5% discount and a yearly fee to maintain the EMR and scheduling service for any vendor that gets more than 10 patients per year from our service. The fee will be 5% of what the vendor generates via our services.

The word Sanjha in Punjabi means a collective or a thing of value that arises only from pooling of sources.( Sehat means Health) 

We are building the digital architecture to support a cooperative of
healthcare stakeholders in Lahore ( and then for other cities in
Pakistan) and enabling pooling of healthcare information that lives in mostly in people's heads and/or hard to navigate and unverified analog and digital
sources at present. 

We are looking for grants and gifts as well as government support to build the registry. Once the registry is made, it will be self-sustaining via intra-app advertisements and starting in year 2, fees from vendors who use our Electronic Medical Record and Scheduling services.

We are applying for grants, talking to wealthy individuals who have an interest in developing healthcare in Pakistan and talking to the Government of Pakistan to share their databases with us.

1. Grants: Solve is the first grant we have applied to but we plan to apply to several more this year and raise at least $100K via grants.

2. We are talking to the head of a major bank in Saudi Arabia. They have had a long-standing interest in optimizing healthcare delivery in Pakistan and will support us if we can meet year 1 milestones of Sehat Sanjha. 

3. Similar conversation with an American entrepreneur of Pakistani origin who lives in San Francisco. 

Both our angel investors would come in at the scale-up phase at the end of year 1.

4. We are in conversation with several branches of Government of Pakistan to see which government healthcare databases - however incomplete - could be used as starters for our registry. This is donation in kind.