What factors actually influence your patient demand?

Today's healthcare landscape of hybrid and virtual care delivery means that patients expect faster access, shorter wait times and a seamless experience wherever they are. As a result, accurately forecasting patient demand is now more important than ever for modern healthcare companies to ensure consistent care delivery without compromising on efficiency.

Below I share some insights from my PhD research at the University of Chicago into the key drivers that actually influence patient demand and show how you can incorporate them into your own forecasts. More practical advice on actually implementing these algorithms is based on my work deploying models to schedule hundreds of thousands of frontline workers at places like Domino’s, Getir, and the Australian Health Department

Why are accurate forecasts important?

Before diving in it’s worth taking a moment to briefly explore why forecasting is so important to workforce analysts and operations teams in healthcare:

1) Short-term staffing optimization: Accurate forecasts ensure proper staffing. Based on our research, 20% of lost utilization comes from staffing at the wrong times — too few staff lead to longer waits and more no-shows, while too many result in wasted resources.

2) Longer-term Capacity Planning: in healthcare bringing additional capacity on board isn't as simple as hiring more staff. Licensing and credentialing processes can extend the ramp-up time for new providers to several months, particularly for clinics operating across multiple states. Accurate long-term forecasting allows you to plan ahead, avoiding the need to resort to expensive agencies during demand spikes.

‍

7 Key Demand Drivers In Healthcare

1. Google Search Trends

At Planbase, we’ve found that including Google search trends as features in our models significantly improves overall performance. These trends are a reliable and interpretable leading indicator of future patient volumes. For example, a surge in searches for ‘‘Ozempic’ in a region often precedes an uptick in clinic visits for diabetes management. By analyzing these search patterns, we can anticipate demand spikes and adjust staffing and resources accordingly.

How to Implement: 

Use tools like Serper to access and store Google Trends data. While this data is retrospective, incorporating simple forecasts based on trend analysis can significantly improve your model's accuracy.

At Planbase we use https://serpapi.com/

2. Seasonality

Seasonality is a fundamental aspect of patient demand forecasting. Many illnesses follow predictable seasonal patterns, directly impacting patient volumes.

How to Implement:

Data science teams can leverage open-source forecasting models like Meta’s Prophet or LightGBM, which have built-in capabilities to handle seasonal patterns. These tools can help you identify and account for recurring yearly trends in your patient data. Often model performance is best when only using data from after the COVID-19 pandemic.

3. Existing Provider Panels

Your current provider panel and care model can significantly influence future demand if your care model necessitates follow-up appointments and ongoing care requirements. If patients are linked to a particular provider it can even be optimal to generate forecasts on a per-provider basis.

How to Implement:

Analyze your care model to identify patterns in follow-up appointments. Integrate this information into your forecasting model to account for the "snowball effect" of initial appointments leading to subsequent visits. 

4. Public Holidays

Public holidays can dramatically affect patient volumes, often leading to spikes before and after the holiday period.

How to Implement: 

Maintain an up-to-date calendar of public holidays across all relevant countries, states, and regions where you operate. Consider using APIs or databases that provide this information to ensure accuracy and ease of maintenance. 

At Planbase we use https://calendarific.com/ 

5. Weather

Weather conditions can significantly impact both brick-and-mortar clinics and virtual urgent care services. It's closely tied to the seasonality of many illnesses and can affect patients' ability or willingness to seek care.

How to Implement: 

Utilize weather data services accessible via API to access real-time and historical weather data. Incorporate this data into your forecasting model to account for weather-related fluctuations in patient demand.

At Planbase we use https://www.visualcrossing.com/

6. School Holidays and Term Dates

School schedules can have a substantial impact on patient volumes, particularly for pediatric services, family-oriented healthcare providers and clinics serving colleges and school groups.

How to Implement: 

Employ web scraping tools to collect school holiday and term date information for the areas you serve. This data can be integrated into your forecasting model to anticipate changes in demand during school breaks and term times.

At Planbase we use https://serper.dev/ and https://openai.com/ 

7. Marketing Spend

Your marketing efforts can directly influence patient demand. Understanding the relationship between marketing spend and patient volumes is crucial for accurate forecasting.

How to Implement: 

Collaborate closely with your marketing team to understand the potential uplift of various campaigns. While automated models can help, manual adjustments based on marketing insights often yield the best results in accounting for marketing-driven demand fluctuations.

At Planbase we recommend users upload their marketing spend to be included when training forecasting models

‍

How Planbase Can Help

By incorporating these seven key drivers into your patient demand forecasting model, you'll be well-equipped to optimize your workforce management strategies. Remember, the goal is not just to react to patient demand but to anticipate and prepare for it proactively.

At Planbase, we understand the complexities of healthcare workforce management in the modern, hybrid healthcare landscape. Our advanced forecasting tools incorporate all these drivers and more, helping healthcare providers optimize their operations, improve patient care, and maintain healthy profit margins.

Ready to take your workforce management to the next level? Contact Planbase today to learn how our SaaS solutions can transform your patient demand forecasting and staffing strategies.

‍