Widespread adoption of Voice over Internet Protocol (VoIP) is underway in the health care industry as hospitals and other facilities across the nation merge their voice and data networks to reduce costs. But as with any new technology, VoIP introduces potential problems. In addition to being a new technology, VoIP is a mission-critical business function. Problems related to voice or video communication can halt hospital operations if your IT staff isn't on top of the situation.

The good news is that most VoIP problems can be easily prevented with a little planning and action before you deploy.

I started as the director of IT for Central DuPage Hospital in Illinois approximately one year ago. The hospital is located in the western suburbs of Chicago with more than 800 dedicated physicians and a staff of 4,000 supporting care centers, occupational health services, and home health and hospice care.

During my time here, I have had the chance to learn firsthand the challenges and lessons of avoiding obstacles through planning prior to implementing new network technologies such as VoIP and wireless. In addition to a significant VoIP deployment, we implemented a wireless telemetry reporting system to provide patient monitoring and automatic notification to staff in the cardiology unit. We also installed a computerized physician order entry system.

Proper tools for the job
When we started with the VoIP implementation, we thought the issue of installation would be easy. But, as we got into the thick of it, we realized we lacked adequate tools to manage the network and VoIP, nor could we troubleshoot connectivity issues that we might experience along the way.

We connected with Jeff Parker, managing partner of OpenWater Solutions, who recommended that our staff invest in an analysis solution that monitors VoIP alongside network and application performance. With VoIP's sensitivity to bandwidth issues and delays, the performance problem of one application can quickly affect the quality of VoIP calls. The right analyzer is critical to accurately troubleshoot VoIP problems and those of other applications.

Beyond accurate troubleshooting, an analyzer helps you to understand how your network is performing and establishes future benchmarks for determining whether VoIP performance is meeting your expectations. After reviewing several options for analyzers, we selected Network Instruments Observer Suite for the project. It was the best deal in terms of providing all of the tools we would need for testing and daily management of VoIP and anything else running on the network. It was also easy to install and use, and reasonably priced compared to other options.

Performing pre-deployment assessments
VoIP isn't the type of technology that network managers should attempt to implement without doing a little legwork up front. A common mistake is to implement now and attempt to troubleshoot issues on the fly. This can result in wasting time working with users to see if problems continue. In turn, management and VoIP users can get burned out.

Conversely, some people anticipate that the addition of VoIP will increase their bandwidth needs and, as a result, automatically upgrade their bandwidth capacity. The increase in bandwidth may waste money without addressing the issue causing performance problems.

Instead of guessing how VoIP would impact our network, Parker used the Observer to conduct a pre-deployment assessment of the network. The assessment prepared our network for the added VoIP traffic, identified and resolved obstacles that would get in the way during implementation, and created performance baselines that we'd use to evaluate post-deployment success. Pre-deployment assessments present a great opportunity to discuss VoIP expectations and adjust for growth. All of this is easier to do prior to deployment, rather than after the fact.

Parker also helped us with implementation of a quality of service solution and verifying the analyzer's ability to give priority to VoIP traffic over other applications. Failing to set precedence for VoIP traffic opens up the possibility of interference from other applications on the network -- a problem known as contention. Contention leads to common VoIP performance problems including jitter and packet loss. Throwing bandwidth at contention problems doesn't solve the issue. Even a network with large bandwidth capacity can have poor call quality due to contention.

Once everything was up and running, we spent time learning what to look for when monitoring VoIP and how to interpret the findings. In addition to asking our users about their experiences with VoIP, we went to our analyzer for the hard numbers. The analyzer gave us multiple metrics on VoIP, as well as an overall call quality score. Call quality scores, such as an R-factor score or a mean opinion score (MOS), can serve as useful VoIP network health indexes, giving you a quick indication of whether a system's performance is improving or degrading.

The R-factor uses a scale of one through 100 to measure call quality. (See figure.) It takes into account how various impairments, including jitter, delay and packet loss, affect the average user's perception of call quality. Typically, the maximum value of a call would be 93.2 after standard degradation. If the R-factor falls below 70, call quality degrades noticeably.

MOS measures user satisfaction of a call. It takes into account a variety of factors impacting call quality, such as handset quality, ambient noise and network performance. MOS measures quality on a scale of one through five, where four or higher indicates a satisfied user. If the average MOS falls below 3.5, it's likely that several VoIP users are dissatisfied. If you notice these statistics trending downward, it's essential to perform more detailed analysis to identify and resolve any performance issues.

Keep in mind that call quality will be perceived differently depending upon the organizatin's environment. For example, a user making a call from a nurse station may tolerate lower quality than a user making a call from a conference room.

The adage "an ounce of prevention is worth a pound of cure" definitely applies to VoIP implementation. The experience of a major VoIP deployment taught us that prevention and planning -- whether in providing care or implementing new technologies -- can make all the difference in avoiding a crisis.

With the right analysis tools in place, knowledge about what we were measuring and planning before installation, we were able to eliminate many of the problems that sometimes trip up installations. Having responsive consultants and the right equipment allowed us to spot issues before they became noticeable to the staff.

Mr. King is IT director for Central DuPage Hospital in Winfield, Ill. He can be reached at (630) 933-1600 or jack_king@cdh.org .