Strengthening Preventative & Curative Care (SPCC)

Multiple diseases are prevalent and remain persistently unaddressed by the public health system in Pakistan. A major reason for this is that preventive and curative care in the primary health system is not targeting the diverse nature of diseases spread across different areas. Primary care policies, resourcing and interventions are developed at a provincial level, and applied across district or local level jurisdictions, ignoring socio-economic contexts or differences in disease patterns and needs within districts. For example, the primary health needs for the province of Khyber Pakhtunkhwa with a population of 40 million is determined centrally based on little evidence. Aggregated and representative data on patient profiling is not measured properly and remains unreliable and sometimes completely unavailable at the district-level.

The District Health Information System (DHIS) data collected at the facility-level in Khyber Pakhtunkhwa (KP) province records patient complaint categories, rather than a full representation of the patient’s record utilizing available medical tools and diagnostic tests. Moreover, the existing staff deployed at these facilities do not have the adequate experience to conduct patient profiling. Thus, the longitudinal view of what is happening in a district or at the sub-district remains unavailable.

The gap in evidence collection and subsequent analysis results in a situation where primary facilities do not have the right medicines, equipment and staff that is most acutely needed in respective districts. Furthermore, in the absence of any sophisticated analysis of the data, timely preventative policies are not devised leaving the system vulnerable.

For instance, during the province-wide Primary Health Care Revamp Initiative Rapid Diagnostic Tests (~70 testing kits across 6 medical conditions[1] per month per facility) and ~90 medicine items (of varying quantities) that were procured for more than 1000 primary facilities were provided using a one size fits all approach rather than any kind of targeting for contexts.

A standardized mechanism to regularly measure and report disease trends at the facility-level will enable public health managers to better plan, resource and deliver preventive and curative health interventions in their respective jurisdictions.

[1] Hepatitis B, Hepatitis C, Pregnancy Test, Blood Sugar, Hemoglobin Testing, Urine Test

We will strengthen the primary healthcare system in KP by achieving the following three objectives:

1. Data collection on software application: Ensure patient’s profile and the diagnosis from a qualified doctor is recorded at the facility in a standardized format and entered into a data base using tablets. The data collection format shall be coded into a software application which will have the functionality of i) recording patient's data; ii) forwarding the patient's information from front desk to the doctor's end for diagnosis and; iii) uploading the patient's attributes with doctor's diagnosis on a central database.
2. Data analysis: Ensure data is analyzed through machine learning models to see district level epidemiological trends, not only providing insights on current disease trends but also predict future disease trends in localities.
3. Data usage: Regional and district level public health managers plan, resource and deliver curative and preventive health interventions based on actionable insights on specific disease trend in their jurisdictions.   

The above-mentioned objectives will be achieved by adopting the following 3 step approach:

1.Data collection - Active screening & early detection

Our solution leverages currently available instruments and equipment available at primary healthcare facilities in KP to conduct a thorough risk assessment of incoming patients. The solution views every patient walking into a primary healthcare facility as a source of valuable information providing insight into the population’s health and epidemiological outcomes. In KP, ~1000 primary healthcare facilities receive ~60,000 patient visits every day.

A digital data collection tool will be developed based on defined metrics and indicators to identify risks among patients visiting primary facilities. The collection of this information will involve building a standard process through which every visiting patient at a primary health facility in KP will go through. This implies that before patients are channeled to the doctor stationed at the clinic, their information will be collected by paramedical staff (also already available in all facilities). Once the patients reach the doctor, their information will already have been passed onto the doctor who in turn will add their diagnosis and prescription.

As an illustrative, few metrics and demographics to be collected are as follows:

1. Age: to determine which illness an individual is susceptible to after reaching a given age threshold
2. Lifestyle habits: pertinent variables like smoking, dipping-tobacco (niswar) and current food intake to determine risk associated with inadequate eating habits and smoking behaviors
3. Weight: to determine the risk of malnourishment/ obesity (associated with broader risks of heart diseases, Type 2 diabetes, stroke and high blood pressure)

Adequate training and clear breakdown of responsibilities will be provided to facility staff to correctly use the data collection tool for incoming patients.

The process flow for facility-level data collection is presented in the chart below:

*These indicators are provided for illustration/example only. The correct indicators will be identified in consultation with health professionals 

There are 23 essential instruments and tools currently available at a Basic Health Unit - a primary healthcare facility - in addition to the 6 RDTs (900+ facilities in KP). As an illustration, data on the following diseases will be collected utilizing the existing instruments and tests:

It must be noted that the data gathered at the facilities will not be limited to the above-mentioned diseases only.

2. Application of big data models for predictive analysis

Once the facility level data has been collected at the facility level, the data will be collated and analyzed at the district-level to assess disease trends and related patient attributes. We intend to build and deploy an algorithm using a standardized repository of evidence/data to understand and predict emerging diseases patterns and possible curative and preventative routes.

In this regard, our model will allow us to achieve the following two objectives

1. Observe diseases prevalent in target facilities and which patient attributes are most associated with the prominent diseases
2. Predict which diseases pose the highest risks to the local population

3.  Data usage

Based on the data collation and analysis, interventions for short, medium and long-term curative and preventative care will be designed at the district level. These include:

• Guide policies and provision of medical and equipment supplies based on identified disease prevalence and data now available at the facility and district-level.

For instance, if in town "Peshawar-I” of district Peshawar, diabetes remains the most prevalent illness among 40+ individuals, the following interventions can be undertaken:

• The predictive analysis would enable public health officials to plan for disease outbreaks and allocate resources and equip facilities in advance.

The solution aims to serve the following:

1.Primary and Secondary Health Department of Khyber Pakhtunkhwa

Our solution is geared towards primarily serving Public Health Managers at the provincial and district-level in the province of KP. The solution will enable them to plan, resource and deliver within the preventative and curative healthcare ecosystem through localized solutions and policies based on data on disease trends.  Our predictive machine learning epidemiology models will enable medium and long-term planning on preventive and curative interventions.

Serving institutional actors can allow for an overall trickle-down effect on the broader community and individuals.

2.Overall community and general population

• Reduced travel time for patients to larger cities and towns due to effective service delivery mechanisms available within local healthcare facilities – contextualized to disease prevalence.
• Timely intervention for identified disease patterns due to availability of relevant medical equipment tools, diagnostic tests, and trained staff.
• Improved service delivery for overall community as patient profiling mechanism will allow for thorough check-ups and increased face time with medical staff, ultimately leading to better patient care.

Impetus Advisory Group is a management consulting firm that provides a range of consulting services including innovation and transformation, strategy, digital and operational performance. The firm is currently supporting critical transformation exercises in health, education and agriculture across the provinces of Khyber Pakhtunkhwa, Sindh, Baluchistan, and the Federal Capital Territory in Pakistan.

Currently, a 14-member team is supporting the Minister of Health and the Health Department of Khyber Pakhtunkhwa undertaking one of the most ambitious transformation of primary and secondary healthcare facilities across the province within the following domains: equipment provision, infrastructure works & medicine availability

Moreover, our team has also supported the Khyber Pakhtunkhwa Government with the following initiatives:

1. Development of Human Resource Management Information System (HRMIS) for the Health Department across the Province
2. Prioritizing current government schemes/projects across Khyber Pakhtunkhwa.

Our support to key government institutions has allowed for our firm to be well-positioned to deliver the following through:

1. Access to key stakeholders within public healthcare – to inform interventions and drive policy
2. Access to ground through field staff support – to understand disease prevalence at local community levels
3. Bridge between healthcare facilities and provincial health secretariat across provincial, district and facility levels. 

The current political landscape of Khyber Pakhtunkhwa remains conducive for large-scale and innovative healthcare initiatives – thus, paving the way for such a solution to be scaled up rapidly.

Our experiences on the field and our support to the Health Department has allowed us to critically identify deficiencies in the primary healthcare system in Khyber Pakhtunkhwa. Currently, there is no mechanism in place that adequately addresses preventative and curative care at the local level based on facility level data. Thus, there is no data to support patient profiling, local-level disease prevalence and risk assessment. We believe we are well situated to solve this problem and ensure scalability.

In addition, our current work with the KP government has allowed us to provide insights directly to provincial health authorities. Therefore, a sound data-driven policy recommendation from our team would overcome the political economy barriers which would otherwise affect the implementation of health-related interventions.

Our solution is innovative and disrupts the healthcare market through

1. Standardize data collection of patient profiles with disease diagnosis: Current data collection regimes at the facility level are inadequate, and don’t provide insights that help public health officials plan and deliver better health outcomes at the local level.

• No data has been collected for the purpose of devising interventions on district-level/ local level disease trends – thus there is an absence of local level actions on preventative and curative diseases.
• Our solution uses existing tools to chalk out analysis to provide actionable insights for healthcare providers and stakeholders and predict disease trends
• The standardized data collection will indirectly improve perception of working of facilities among population and will simultaneously serve as a quality of care improvement intervention

2. Usage of existing resources:

• Our solution leverages the existing human resource to collect data.
• Our solution uses existing instrument and equipment to conduct risk assessment and formulate analysis

3. Implementation of modern machine learning techniques: In KP, facility level data is sometimes used to report descriptive statistics and no sophisticated analysis is employed

• We will use patient profile and disease diagnosis to provide data-backed solutions to not only allocate current resources based on disease severity but also plan, allocate and procure resources in the medium to long-term based on predicting future disease patterns.

Our impact goals with the plan for next year and next 5 years are presented in the image below:

Our impact goals will be determined by our ability to guide health policies and resource management at the district level. Our impact goals are further aggregated into short, medium and long-term goals.

Short-term goal: Performance management check for facility-staff reflected in their ability to conduct initial patient assessment and collect patient data – to measure service delivery and quality of care

Medium- term goal: Epidemiological models to predict diseases patterns among communities and stakeholders to plan, resource and implement interventions accordingly

Long-term goal: Overall reduction in disease prevalence for the measured diseases

An aspect of our solution is to empower supervisors of primary healthcare facilities, usually a Medical Officer (MO), by providing them with adequate training on

1)Using all the existing medical equipment and rapid diagnostic test (RDTs) on incoming patients

2)Recording their data on the software provided by the project team

Therefore, these trained MOs will record patients’ data and add their profile in the system for the project team to analyze and predict 

We aim to achieve a gender balance by having at least 50% female staff in our team. Moreover, our solution is currently geared towards Peshawar district (for potential piloting phase), whereby the gender distribution of incoming patients remains similar. Moreover, being the capital of the Khyber Pakhtunkhwa Province, the district welcomes patients from across the province – including marginalized communities, various tribes, ethnic groups and individuals of all ages. 

• Being a growing consulting and professional services firm, we would only be looking for seed funding for this concept. Once designed and piloted, ownership of devised solution will be transferred and adopted by the Government of Khyber Pakhtunkhwa
• Our current proximity to the Government of Khyber Pakhtunkhwa enables us to pilot, scale and ensure complete adoption of the solution by health managers at the district and provincial level.