By A.G. Breitenstein
Growing accountability and financial penalties placed on hospitals for potentially preventable hospital-acquired complications - such as infections, pulmonary emboli, venous thrombosis and decubitus ulcers - have placed hospitals' bottom lines at risk. Potentially preventable complications can add 10 percent to national hospital inpatient costs, or $88 billion a year. If preventable complications were reduced or eliminated, the savings would significantly impact the financial performance of the country's healthcare system.
Timely and accurate patient-level information should be delivered to care providers while the patient is in the care setting. Essential patient data often resides in disparate information systems, not in the hands of a clinician during a patient encounter. Compounding this issue, few hospitals have sufficient staff members with the analytic expertise needed to integrate the data into real-time medical practice. Even in cases where they have both the data and expertise, few hospitals have enough comparative data to understand their medical activities in proper statistical context.
Many hospitals lack a systems-based approach to identify patients at risk for preventable complications. These same hospitals are often unable to track whether or not the appropriate steps have been taken to address that risk and improve the quality of care. The results can be dangerous and costly. In a recent analysis, nine studied preventable complications - including decubitus ulcers, post-op VTE, post-op respiratory failure and MRSA - resulted in nearly 7,000 excess length-of-stay days and added approximately $23 million to hospital expenses for an average 250-bed hospital.
A real-time predictive analytic solution that provides the following capabilities and features can have a positive impact:
- real-time and integrated data from disparate hospital IT systems;
- advanced data-driven predictive analytics and modeling techniques that help identify high-risk and high-cost patients;
- real-time alerts to physicians and other caregivers;
- identification of appropriate prophylactic processes and procedures;
- necessary care tracking and support for clinicians as they intervene to improve care; and
- generation of retrospective analytic reports that measure the characteristics and effect of the response taken.
Imagine a hospital implements an initiative to reduce preventable bleeding for those patients on anticoagulation medication, such as warfarin. A real-time, predictive analytic system could detect all patients on anticoagulation medication and identify those patients who are at greatest risk for bleeding. This prediction model takes into account the patient's inherent risk (i.e., patient and family history) and current medical condition, based on INR and PTT test values, blood pressure, renal function and the patient's comorbidities to determine the bleeding risk. Accompanying software monitors the patient in real time as new data becomes available, such as new lab results and newly initiated medications. Based on this information, the software tool alerts the relevant personnel to the high-risk patients.
The software also tracks and provides updates on adherence with evidence-based protocols and Joint Commission standards. Finally, the tool provides identification of pertinent patient information (e.g., specific floor, facility and attending physician), documentation of procedures, manual over-writing capabilities and corresponding outcomes. These analyses will facilitate identification and, ultimately, reduction of treatment variation. The likely result is a significant reduction in unreimbursed preventable complications, length of stay, and patient morbidity and mortality.
Choosing the right solution can be daunting and expensive. Hospitals should invest in a robust predictive-analytic solution that is both capital efficient and easily implemented in order to avoid putting unnecessary strain on their IT staff and bottom line. In addition, hospitals should look for the following features as they search for the right predictive-analytics solution to help improve patient care:
- Software-as-a-service (SaaS) delivery: SaaS-based technologies require minimal resources to install and maintain.
- Compliant with national standards: Technology can help ensure compliance with The Joint Commission's core measures, as well as other national and organizational standards, helping organizations focus more on patient care and less on regulations.
- Privacy and security: Any technology implemented should comply with HIPAA regulations and protect all patient-level data at the highest level.
About the author
A.G. Breitenstein is vice president and general manager of provider markets for Humedica. For more information on Humedica solutions, click here.