Presentation
PS7 - Investigating the Impact of Healthcare Restructuring on Patient Experience through Data Mining
DescriptionBackground:
Healthcare organizations commonly make decisions to restructure their business, which results in hospital closures, mergers (in which two organizations merge to form one organization), and acquisitions (in which one organization purchases another and incorporates it into their system). These decisions can significantly impact healthcare access, quality of care, patient experience, patient outcomes, and healthcare associated costs for patients. (Barker et al., 2022; Helfand et al., 2013; Ho & Hamilton, 2000; Mariani et al., 2022; Su, 2017)
While some studies suggest that hospital consolidation can enhance care quality, others indicate worsened clinical outcomes (Barker et al., 2022; Mariani et al., 2022). Mergers, acquisitions, and closures alter healthcare service availability. causing longer wait times, delayed diagnoses, or inadequate treatment, and often disproportionately affecting vulnerable populations (Ho & Hamilton, 2000; Kaufman et al., 2016). In particular, closures, in rural and underserved areas, result in longer travel times, reduced access to specialized services, and increased wait times for appointments and procedures, which has direct effects on patient health outcomes, particularly for patients with chronic conditions who require ongoing care (Gujral & Basu, 2019; Helfand et al., 2013; Kaufman et al., 2016). Closures also cause out-migration of highly trained healthcare professionals, which can force patients to seek care from providers who are less experienced or have fewer resources (Gujral & Basu, 2019).
Importantly, hospital restructuring is a result of top-down organizational decision making that prioritizes the health of the business rather than the health of the community (Barker et al., 2022; Kaufman et al., 2016). Understanding the potential cumulative impacts that these business decisions can have on access and quality of care is crucial to helping healthcare organizations and healthcare providers develop strategies to maintain high quality care and mitigate the potential harms to patients and communities (Barker et al., 2022; Mariani et al., 2022; Panagiotoglou et al., 2016; Su, 2017).
Research Objectives:
By studying the impact of healthcare restructuring on the quality of healthcare provided, we can identify common challenges and potential areas for improvement. Here, we discuss our preliminary findings on the impacts of restructuring on patient experience. We are particularly interested in understanding these impacts due to an upcoming change in healthcare access in our region related to one healthcare system acquiring another.
Methodology:
We used Python and the Selenium package to scrape Becker’s Hospital Review newsletters posted between January 2020 and October 2023 to identify hospital restructuring events across the United States, including mergers, acquisitions, and closures. Keywords of “close”, “merge”, and “acquire”, including common permutations of these words (e.g., “close” and “closing”) contained within the titles of articles included within Becker newsletters were automatically flagged in our data using python. Next, to compare patient experience prior to and after a restructure, we used web scraping and Large Language Models (LLMs) to automatically identify the exact hospital and healthcare center locations involved in mergers or acquisitions, and the dates at which these changes took effect. We then used Google and Yelp Reviews posted in a six-month window prior to, and six months after the date the restructure took place to examine the impact of the restructure on patient experience. Current results include a description and evaluation of the methods we used to generate a large dataset through web scraping and the application of LLMs, as well as quantitative analysis of the average star ratings (i.e., how many stars were given out of 5 stars).
Preliminary Results:
Our technique of web scraping healthcare newsletters and applying LLMs to create a large dataset of healthcare business restructuring events demonstrated potential for automating a large part of the manual data search and cleaning process. Data extracted from the Becker Hospital Review newsletters between January 1 2020 and October 14th 2023 included a total of 14,530 articles. Of these, titles that included keywords or common permutations included 294 articles that mention "merge", 114 that mention "acquire", and 478 that mention "closure". We then used web scraping and LLMs to identify the organizations and dates of business restructuring. Initially, we leveraged the OpenAI GPT-3.5 API, using specific queries to extract names, locations, and merger dates. However, the accuracy overall in was low and inconsistent. To streamline the process, we used Python and Selenium to automate interactions with ChatGPT by programmatically entering prompts onto the website and recording responses from ChatGPT into our dataset. While this approach streamlined the process, it was not 100% accurate, and therefore manual review was still necessary to confirm that the correct hospitals and dates were captured.
From the data processed, we randomly selected 10 examples of healthcare centers that experienced a merger or acquisition and manually collected Google and Yelp Review data of these systems in a six-month pre-change and six-month post-change period. To compare how these restructure types and pre and post-change periods impacted patient experience review scores, we conducted a two-way ANOVA with restructure type and time period as IVs, and patient experience review scores as the DV. Results indicated significant main effects of restructure type (F(1, 648) = 5.51, p = 0.019) and time period (F(1, 648) = 11.62, p < 0.001), and a non-significant interaction (F(1, 648) = 2.52, p = 0.113). Average patient experience review star ratings were lower in the post-period than the pre-period for both acquisitions and mergers. While average patient experience star ratings for hospitals involved in an acquisition were higher in both pre and post-periods than ratings for hospitals that experienced a merger, the drop in average star ratings in the post-period was more pronounced.
Preliminary Conclusion:
These initial findings shed some light on the influence that business decisions can have on the experiences of patients as they seek care from organizations that undergo restructuring. While analysis is ongoing, this research demonstrates that data mining is a valuable tool in examining national trends and predicting the impacts similar changes can have on regions as organizations change the way care is delivered to communities. Future work will focus on recognizing how organizations can prepare for changes in community access to various care services (e.g., mental and maternal health) to ensure that local populations do not experience major disruptions to their treatment.
References:
Barker, E. H., Watt, J., Tranmer, J., Clark, R., Town, R., Hickman, B., Gowrisankaran, G., Gil, R., Lange, F., Arslan, A., Lehrer, S., Gregory, A. W., & Dimitropoulos, D. (2022). The Impact of Hospital Closures and Mergers on Patient Welfare *. https://doi.org/10.1787/8d805ea1-en
Gujral, K., & Basu, A. (2019). NBER WORKING PAPER SERIES IMPACT OF RURAL AND URBAN HOSPITAL CLOSURES ON INPATIENT MORTALITY. http://www.nber.org/papers/w26182
Helfand, M., Peterson, K., & Humphrey, L. (2013). Evidence-based Synthesis Program Evidence Brief: Effects of Small Hospital Closure on Patient Health Outcomes.
Ho, V., & Hamilton, B. H. (2000). Hospital mergers and acquisitions: does market consolidation harm patients? In Journal of Health Economics (Vol. 19). www.elsevier.nlrlocatereconbase
Kaufman, B. G., Thomas, S. R., Randolph, R. K., Perry, J. R., Thompson, K. W., Holmes, G. M., & Pink, G. H. (2016). The Rising Rate of Rural Hospital Closures. Journal of Rural Health, 32(1), 35–43. https://doi.org/10.1111/jrh.12128
Mariani, M., Sisti, L. G., Isonne, C., Nardi, A., Mete, R., Ricciardi, W., Villari, P., De Vito, C., & Damiani, G. (2022). Impact of hospital mergers: a systematic review focusing on healthcare quality measures. European Journal of Public Health, 32(2), 191–199. https://doi.org/10.1093/eurpub/ckac002
Panagiotoglou, D., Law, M. R., & Mcgrail, K. (2016). Effect of Hospital Closures on Acute Care Outcomes in British Columbia, Canada An Interrupted Time Series Study. www.lww-medicalcare.com
Su, E. (2017). Hospital merger and acquisition effects on healthcare quality and cost.
Healthcare organizations commonly make decisions to restructure their business, which results in hospital closures, mergers (in which two organizations merge to form one organization), and acquisitions (in which one organization purchases another and incorporates it into their system). These decisions can significantly impact healthcare access, quality of care, patient experience, patient outcomes, and healthcare associated costs for patients. (Barker et al., 2022; Helfand et al., 2013; Ho & Hamilton, 2000; Mariani et al., 2022; Su, 2017)
While some studies suggest that hospital consolidation can enhance care quality, others indicate worsened clinical outcomes (Barker et al., 2022; Mariani et al., 2022). Mergers, acquisitions, and closures alter healthcare service availability. causing longer wait times, delayed diagnoses, or inadequate treatment, and often disproportionately affecting vulnerable populations (Ho & Hamilton, 2000; Kaufman et al., 2016). In particular, closures, in rural and underserved areas, result in longer travel times, reduced access to specialized services, and increased wait times for appointments and procedures, which has direct effects on patient health outcomes, particularly for patients with chronic conditions who require ongoing care (Gujral & Basu, 2019; Helfand et al., 2013; Kaufman et al., 2016). Closures also cause out-migration of highly trained healthcare professionals, which can force patients to seek care from providers who are less experienced or have fewer resources (Gujral & Basu, 2019).
Importantly, hospital restructuring is a result of top-down organizational decision making that prioritizes the health of the business rather than the health of the community (Barker et al., 2022; Kaufman et al., 2016). Understanding the potential cumulative impacts that these business decisions can have on access and quality of care is crucial to helping healthcare organizations and healthcare providers develop strategies to maintain high quality care and mitigate the potential harms to patients and communities (Barker et al., 2022; Mariani et al., 2022; Panagiotoglou et al., 2016; Su, 2017).
Research Objectives:
By studying the impact of healthcare restructuring on the quality of healthcare provided, we can identify common challenges and potential areas for improvement. Here, we discuss our preliminary findings on the impacts of restructuring on patient experience. We are particularly interested in understanding these impacts due to an upcoming change in healthcare access in our region related to one healthcare system acquiring another.
Methodology:
We used Python and the Selenium package to scrape Becker’s Hospital Review newsletters posted between January 2020 and October 2023 to identify hospital restructuring events across the United States, including mergers, acquisitions, and closures. Keywords of “close”, “merge”, and “acquire”, including common permutations of these words (e.g., “close” and “closing”) contained within the titles of articles included within Becker newsletters were automatically flagged in our data using python. Next, to compare patient experience prior to and after a restructure, we used web scraping and Large Language Models (LLMs) to automatically identify the exact hospital and healthcare center locations involved in mergers or acquisitions, and the dates at which these changes took effect. We then used Google and Yelp Reviews posted in a six-month window prior to, and six months after the date the restructure took place to examine the impact of the restructure on patient experience. Current results include a description and evaluation of the methods we used to generate a large dataset through web scraping and the application of LLMs, as well as quantitative analysis of the average star ratings (i.e., how many stars were given out of 5 stars).
Preliminary Results:
Our technique of web scraping healthcare newsletters and applying LLMs to create a large dataset of healthcare business restructuring events demonstrated potential for automating a large part of the manual data search and cleaning process. Data extracted from the Becker Hospital Review newsletters between January 1 2020 and October 14th 2023 included a total of 14,530 articles. Of these, titles that included keywords or common permutations included 294 articles that mention "merge", 114 that mention "acquire", and 478 that mention "closure". We then used web scraping and LLMs to identify the organizations and dates of business restructuring. Initially, we leveraged the OpenAI GPT-3.5 API, using specific queries to extract names, locations, and merger dates. However, the accuracy overall in was low and inconsistent. To streamline the process, we used Python and Selenium to automate interactions with ChatGPT by programmatically entering prompts onto the website and recording responses from ChatGPT into our dataset. While this approach streamlined the process, it was not 100% accurate, and therefore manual review was still necessary to confirm that the correct hospitals and dates were captured.
From the data processed, we randomly selected 10 examples of healthcare centers that experienced a merger or acquisition and manually collected Google and Yelp Review data of these systems in a six-month pre-change and six-month post-change period. To compare how these restructure types and pre and post-change periods impacted patient experience review scores, we conducted a two-way ANOVA with restructure type and time period as IVs, and patient experience review scores as the DV. Results indicated significant main effects of restructure type (F(1, 648) = 5.51, p = 0.019) and time period (F(1, 648) = 11.62, p < 0.001), and a non-significant interaction (F(1, 648) = 2.52, p = 0.113). Average patient experience review star ratings were lower in the post-period than the pre-period for both acquisitions and mergers. While average patient experience star ratings for hospitals involved in an acquisition were higher in both pre and post-periods than ratings for hospitals that experienced a merger, the drop in average star ratings in the post-period was more pronounced.
Preliminary Conclusion:
These initial findings shed some light on the influence that business decisions can have on the experiences of patients as they seek care from organizations that undergo restructuring. While analysis is ongoing, this research demonstrates that data mining is a valuable tool in examining national trends and predicting the impacts similar changes can have on regions as organizations change the way care is delivered to communities. Future work will focus on recognizing how organizations can prepare for changes in community access to various care services (e.g., mental and maternal health) to ensure that local populations do not experience major disruptions to their treatment.
References:
Barker, E. H., Watt, J., Tranmer, J., Clark, R., Town, R., Hickman, B., Gowrisankaran, G., Gil, R., Lange, F., Arslan, A., Lehrer, S., Gregory, A. W., & Dimitropoulos, D. (2022). The Impact of Hospital Closures and Mergers on Patient Welfare *. https://doi.org/10.1787/8d805ea1-en
Gujral, K., & Basu, A. (2019). NBER WORKING PAPER SERIES IMPACT OF RURAL AND URBAN HOSPITAL CLOSURES ON INPATIENT MORTALITY. http://www.nber.org/papers/w26182
Helfand, M., Peterson, K., & Humphrey, L. (2013). Evidence-based Synthesis Program Evidence Brief: Effects of Small Hospital Closure on Patient Health Outcomes.
Ho, V., & Hamilton, B. H. (2000). Hospital mergers and acquisitions: does market consolidation harm patients? In Journal of Health Economics (Vol. 19). www.elsevier.nlrlocatereconbase
Kaufman, B. G., Thomas, S. R., Randolph, R. K., Perry, J. R., Thompson, K. W., Holmes, G. M., & Pink, G. H. (2016). The Rising Rate of Rural Hospital Closures. Journal of Rural Health, 32(1), 35–43. https://doi.org/10.1111/jrh.12128
Mariani, M., Sisti, L. G., Isonne, C., Nardi, A., Mete, R., Ricciardi, W., Villari, P., De Vito, C., & Damiani, G. (2022). Impact of hospital mergers: a systematic review focusing on healthcare quality measures. European Journal of Public Health, 32(2), 191–199. https://doi.org/10.1093/eurpub/ckac002
Panagiotoglou, D., Law, M. R., & Mcgrail, K. (2016). Effect of Hospital Closures on Acute Care Outcomes in British Columbia, Canada An Interrupted Time Series Study. www.lww-medicalcare.com
Su, E. (2017). Hospital merger and acquisition effects on healthcare quality and cost.
Event Type
Poster Presentation
TimeTuesday, March 264:45pm - 6:15pm CDT
LocationSalon C
Digital Health
Simulation and Education
Hospital Environments
Medical and Drug Delivery Devices
Patient Safety Research and Initiatives