New technology lowers LOS, saves $2.5 million

By David Dillehunt, CSP

Chief Information Officer

Cape Fear Valley Health System

Fayetteville, NC

Health care facilities and integrated delivery networks are continually challenged by heightened competition for patients and physicians, demands to decrease operating costs, and pressure to increase efficiency. One widespread solution to these challenges has come in the form of information systems technology.

Cape Fear Valley Health System (CFVHS) is a large, integrated health care delivery network comprising four hospitals, 16 primary care physician practices, and six specialty physician practices. It has experience with many of the obstacles common to health care organizations in the modern managed care environment.

To address these issues at Cape Fear, George Binder, MD, chief of radiology services, and I began to research clinical image management systems (CIMS), also known as PACS.

Our goal was to identify a CIMS that would provide two basic benefits to the health system:

1. We wanted a system that could provide superior access to clinical images and corresponding diagnostic information on an enterprisewide level. We decided that to have a tangible impact on the efficiency of the entire health system, our prospective CIMS must extend well beyond the borders of the radiology department and allow information sharing throughout the organization.

2. We wanted to sharply reduce film and film-related expenses within the radiology department. We associated a significant savings with the reduction of these expenses, although our primary concern was the efficiency of the overall enterprise.

The system we selected was ImageACCESS, a Windows NT-based CIMS from StorCOMM in Jacksonville, FL.

How the technology works

The CIMS performs four major categories of functionality:

• Acquisition functions capture images for use in the system.

• Storage functions store and file electronic images.

• Retrieval and distribution functions move images to the end user.

• Display functions allow the user to review images.

At CFVHS, the CIMS acquires images from CT, magnetic resonance (MR), ultrasound, and plain film via film scanners. The CT and MR devices at the medical center are DICOM-compliant modalities. DICOM (digital imaging and communications in medicine) is a standard protocol that formats image information at the device level.

The CIMS can use raw and processed DICOM data directly imported from those devices. The CIMS acquires images from the non-DICOM-compliant devices via its image capture software. Both DICOM and the image capture software acquisition interfaces result in an electronic collection of data that contains all the clinical images in the exam and associated patient information. Users also can add reports, annotations, or comments.

Electronic exams are stored on a centralized electronic database. The image server is a traffic cop for the entire system, processing requests and retrieving/distributing exams on demand. Images are processed and uncompressed in two phases.

The system initially stores image information on RAID (redundant array of inexpensive drives), a series of high-capacity hard drives allowing fast access speeds. After seven days on RAID, exams are transferred to less expensive archive storage media. We chose a seven-day active period since demand for exams is considerably less after one week. We currently use two types of archive storage: WORM drives (write once, read many) and a magneto optical jukebox. The two-phase storage strategy is a compromise between slightly slower access speeds and less-expensive storage media.

The CIMS retrieval and distribution mechanisms are built on an open client/server architecture. Display stations located throughout the medical center run the CIMS’ view station software. We currently have stations in radiology, emergency medicine, the operating room, and five intensive care units — surgical, medical, coronary, neonatal, and cardiac surgery.

The view station software can be customized based on user name, so that no matter where a user logs on to the software, he or she will encounter the same interface.

As I watched the impact of the CIMS on the radiology department, it became apparent that other areas, specifically the intensive care units, also were benefiting from the new technology. I decided to conduct a study measuring the effectiveness of the CIMS in other areas of the medical center.

As decreased patient length of stay (LOS) is a good indicator of increased efficiency and revenue opportunities and of dollars saved by the facility, I designed the study to compare the average LOS in three intensive care units — in two similar pre- and post-implementation periods. I theorized that real-time access to clinical images and related diagnostic information would contribute to a reduction in LOS.

I audited the LOS for 1,306 patients who had been admitted to the surgical, medical, or coronary intensive care units. Approximately 632 patients had been admitted during a pre-implementation span from December 1997 to February 1998, and 674 had been admitted during the post-implementation span from December 1998 to February 1999.

After the results had been analyzed, we found that a decrease of .75 days had occurred from a 10.09 day average in the pre-implementation period to a 9.34 day average in the post-implementation period. We also proved, with a 95% statistical confidence level, that this reduction was not a coincidence.

Even more encouraging were the results we obtained by applying cost-accounting methodology to the LOS numbers. Using this method, I calculated a drop in total direct costs for those patients treated, between the two three-month periods analyzed, that showed a savings of $647,000, or $2.5 million per year.

Our methodology measured all direct costs that comprise patient care, such as unit staffing, actual procedure costs, and supplies. Indirect (overhead) costs such as electricity and administration were not taken into account. That was accomplished at a time when the case mix index actually rose by 6.8%, meaning that the patients were actually sicker.

A previous study, conducted by Carmen Perri, an ultrasound supervisor at CFVHS, also demonstrated additional savings to the health system resulting from the use of the CIMS. This study showed that the savings and increase in productivity brought about by the CIMS were projected to increase revenue in the ultrasound department alone by nearly $1.2 million per year.

Combining the results of these two studies produces an annualized benefit of $3.7 million — $1.2 million in increased revenue and $2.5 million in actual reduced costs.

That figure does not take into account areas where savings are presumed but not yet quantified, such as the radiology department in general or the emergency medicine department. In any case, $3.7 million represents a phenomenal return, given an overall investment of approximately $2.3 million in the CIMS system.

The future of the technology

There is no question that information systems are a necessity in the modern health care industry. Electronic clinical image management has allowed us to share information on demand throughout the health system; improve work flow, and increase our efficiency; decrease our operating costs and increase our revenue enhancement opportunities; and enable us to compete for physicians and provide the highest caliber of patient care.

CIMS is the next generation of technology, allowing health care organizations to apply the benefits of information systems to clinical images. In my opinion, CIMS will soon be indispensable to the health care enterprise.