--Put your SQL code in this block
SELECT COUNT (*) AS Bloodproduct_count, ordercategoryname, caregivers.cgid, admission_type
FROM caregivers
INNER JOIN inputevents_mv
ON caregivers.cgid = inputevents_mv.cgid
INNER JOIN admissions
ON inputevents_mv.hadm_id = admissions.hadm_id
WHERE INPUTEVENTS_MV.ordercategoryname LIKE '%blood products%'
Group by ordercategoryname, caregivers.cgid
HAVING Bloodproduct_count > 3Analysis Report Three - Information Systems and Healthcare Provider Interactions
Executive Summary
Healthcare is leaning on technology increasingly, and it continues to evolve and cause frustration among healthcare professionals. When implementing changes to workflows within EHRs, it is important to communicate effectively, set realistic expectations, and provide your staff with training and support. AI continues to evolve to help identify issues early, but it’s important that it not be used solely and that it be used alongside a healthcare professional. Overall, technology can improve efficiency and patient care, but it also requires continuous training as technology and healthcare evolve.
Introduction
“The Journey Through Grief” article from this week’s reading discusses the challenges healthcare staff and the organization itself encounter when implementing a new EHR. Many healthcare staff do not want to evolve with technology or learn new workflows and want to keep things the same because change can cause frustration which leads to resistance. The article explains the challenges thorough the stages of grief, denial, anger, bargaining, depression and finally acceptance. Give staff realistic expectations, and let them know that the organization there to support them with training or any assistance they may need in the process. (McAlearney et al. 2015)
“Seven Years after Meaningful Use”, identifies benefits and challenges nurse and physicians have encountered when using EHRs. Benefits included access to real-time data, helpful alerts, and ease of sharing with other staff. Challenges included, increased workload, and losing access to the EHR. Technology and healthcare is continuing to change, “Static solutions are likely to fail, so practitioners and policymakers much search for ways to adapt and help the health system effectively adapt and evolve.” (Rathert et al. 2019)
“When AI overrules the nurses caring for you” discusses how AI is now using data to help identify issues early. While this can be helpful, it also identifies there are issues with AI and does not always consider the “whole” picture of the patient. For example, a sepsis alert was identified for a patient, and AI did not take into account the patient had leukemia, a diagnosis that could cause similar blood counts. AI works best together with a healthcare professional, while AI helps guide decisions it should never replace clinical judgement. (Bannon 2023)
The Healthcare Context
“(Blood) transfusions can lead to serious harm if safety protocols are not strictly followed. Errors—ranging from mismatched blood types to administrative oversights—can cause severe complications, including hemolytic reactions and organ failure” (Dash, Al-Snan, and Messaoudi 2026). Electronic protocols are set in place to help prevent errors and patient harm. In transfusion medicine, AI is being utilized to reduce waste and make clinical decisions faster by using patient parameters and inventory management. AI is not “bullet proof” and continues to have obstacles, for example transfusion prediction is based on past data and does not consider things such as seasonal variation. There are also obstacles involving AI accurately predicting transfusion reactions. (Arcot et al. 2026) Overall, AI can help improve efficiency and early detection but should not replace a healthcare professionals clinical judgement.
Data Visualizations
Visualization One
#Put your ggplot visualization in this block
ggplot(data = myquery1,
aes(x = cgid, y = Bloodproduct_count, fill = admission_type)) +
geom_col() +
labs(title = "Plot showing blood products given by Caregiver and Admission Type")
This diagram joined the caregivers, inputevents_mv and admissions table. The caregivers table was needed to identify the caregiver ID, inputevent_mv was needed to be able to identify blood products given, and the admissions table was needed to determine admission type. I also included the COUNT statement to include the count of blood_products given and used HAVING to include blood products > 3. This data was put together to visualize the association of blood products given by individual caregivers and looked to determine blood_products given in the Emergency setting vs urgent setting. Looking at the data, blood products were primarily given in the emergency dept compared to the urgent care setting, this is likely due to the fact that the emergency setting where they encounter trauma-related cases, like car accidents, but would need to be looked into more to determine.
Visualization Two
--Put your SQL code in this block
SELECT COUNT(*) AS blood_products, caregivers.cgid, diagnosis
FROM caregivers
INNER JOIN inputevents_mv
ON caregivers.cgid = inputevents_mv.cgid
INNER JOIN admissions
ON inputevents_mv.hadm_id = admissions.hadm_id
and caregivers.cgid IN (15564, 16890, 17457, 21050)
Group by caregivers.cgid, diagnosis
HAVING blood_products > 50 #Put your ggplot visualization in this block
ggplot(data = myquery2,
aes(x=cgid, y=blood_products, fill=diagnosis))+
geom_col() +
labs(
title = "Amt of Blood Products given by Caregiver ID related to diagnosis",
x="Caregiver ID",
y="Number of Blood Products"
)
This diagram joined the caregivers, inputevents_mv and admissions table. The caregivers table was needed to identify the caregiver ID, inputevent_mv was needed to be able to identify blood products given, and the admissions table was needed to determine the diagnosis. I also included the COUNT statement to include the count of blood_products given and used HAVING to include blood products > 50. This data was put together to determine if there any any correlation with diagnosis and blood transfusions as well as see if there was any correlation with caregiver ID’s. I would have to look more into the data, but it is likely that these caregivers may work on different floors or specialties due to the fact that none have any matching diagnosis. For example, caregiver 17457 may work on the GI floor because they have only given blood products to patients with a diagnosis of a GI bleed.
Recommendations for Industry
With blood products being primarily used in the emergency setting, I recommend reviewing the data to determine if there are frequent ED utilizers. For example, Mr. Smith is admitted to the ED every few months for low HGB and frequently requires transfusions. As administrators, we need to look deeper, social determinants of health may need to be evaluated to determine the reason for this. “Socioeconomic status (SES) has been identified as a fundamental cause of disease. People who are poor and powerless have worse health and longevity than those with money, power, and prestige.” (Flaskerud, DeLilly, and Flaskerud 2012) Does Mr. Smith have transportation issues and can’t get to and from appointments or does he have high medical costs or no insurance? If these are identified, a social worker needs to get involved to help find access to transportation and/or assistant with healthcare coverage or cost. This not only improves patient care but decreases utilization. Once this is identified, the patient should be enrolled in a ED-based case management program (Walter et al. 2021) and a flag should be placed in the EHR to easily identify for future ED visits. This should be applied to all frequent utilizers of the ED, not just ones receiving blood products.
References
Arcot, Priyadarsini Jayachandran, Najla Haneefa Basheela, Suvro Sankha Datta, Karan Kumar, and Suchet Sachdev. 2026. “AI in Transfusion Medicine: A Scoping Review with Bibliography.” Indian Journal of Hematology and Blood Transfusion, 1–11.
Bannon, Lisa. 2023. “When AI Overrules the Nurses Caring for You.” The Wall Street Journal 15.
Dash, Hirak Ranjan, Noora Rashid Al-Snan, and Safia Abdessalem Messaoudi. 2026. “Forensic Investigation of Blood Transfusion: Medical Errors, Detection, and Legal Accountability.” In Forensic Investigation of Blood Evidence, 103–21. Springer.
McAlearney, Ann Scheck, Jennifer L Hefner, Cynthia J Sieck, and Timothy R Huerta. 2015. “The Journey Through Grief: Insights from a Qualitative Study of Electronic Health Record Implementation.” Health Services Research 50 (2): 462–88.
Rathert, Cheryl, Tracy H Porter, Jessica N Mittler, and Michelle Fleig-Palmer. 2019. “Seven Years After Meaningful Use: Physicians’ and Nurses’ Experiences with Electronic Health Records.” Health Care Management Review 44 (1): 30–40.
Walter, Lauren A, Elizabeth M Schoenfeld, Catherine H Smith, Erin Shufflebarger, Charles Khoury, Katherine Baldwin, Jennifer Hess, et al. 2021. “Emergency Department–Based Interventions Affecting Social Determinants of Health in the United States: A Scoping Review.” Academic Emergency Medicine 28 (6): 666–74.