I spoke with founder and managing director of Spyglass Consulting,
Gregg Malkary yesterday about two of his most recent studies. The latest is
Point of Care Computing for Nursing (press
release - pdf), and released this October, the same topic targeting physicians (press
release - pdf). Gregg did his first study on mobile computing
in 2003, with follow on studies targeting nurses and physicians in 2004
and 2005. Two years down the road, it was time to take another look -
with some surprising results. The bottom line: significant progress in many areas, but overall there's still a long way to go.
This latest study provides a nationwide perspective that reveals both product
shortcomings and pitfalls for deploying technologies. In further support of the adage, "you don't know what you
don't know," some of the findings are counter intuitive. According to
Gregg, a lot of the findings are actually misconceptions. Impressions created by trade publications and vendor sales and marketing creates most of these
misconceptions - held by providers and vendors who have drunk their own (or others) Kool-Aide. Let's start with
The study found that 64% of nurses in more than
half of the hospitals
believe their wireless network is not reliable enough to support point
of care computing solutions. That's a frightening majority of the hospitals with poorly
designed, installed and/or managed wireless LANs. This contrasts with
the perception that wireless LANs are pervasively deployed, with great
coverage and solid performance. The study describes a
hospital in the Chicago area where nurses have taped X's on the floor
were they can get a good network connection. With the
exception of the leading edge hospitals - the magnet or "most
wired" hospitals - wireless network deployments in patient care areas
have a long way to go.
Another misconception includes
computers on wheels (COWs) that are intended to be rolled up to the bedside for meds
administration and paperless charting. Barcoded patient
wristbands are read by the scanner mounted on the COW. The
survey found that the majority of COWs remain in the hallway and
are rarely (if ever) pushed to the bedside. And those barcodes are
frequently hard to scan (especially depending on the selection of
barcode, printers and wristband material). Hospitals were found to make extra
copies of wristband barcodes (or to cut them off patient wrists) and scanned at the nursing
station - apparently where it's more convenient to scan the darn things
several times each. Perhaps if the cable on the barcode scanner were
long enough to reach the patient from the hallway...
One of the conclusions from the report is that many point of care devices are not well integrated
with applications and workflow. Many applications are not tuned for
clinicians and are hard to use. There are few examples where point of care device and application vendors get together to, you know, make their stuff work well together.
Most point of care devices (with the exception of Motion's C5, the Emano Tec, and a very few others) are really intended for corporate applications, your stock broker or the UPS gal. These devices lack key features for clinical environments. And it seems that neither device or software vendors want to optimize applications for specific devices.
Security was another common complaint. The
way many IT departments have implemented security it takes 1 to 3
minutes to log in. And of course every time the caregiver has to leave
the keyboard - and they're constantly interrupted - they have to log
back in. Many caregivers end up making notes on paper and then typing them into the EMR later in the shift when they have more time.
The integration of technologies, training, and user acceptance were found to be key to successful deployments. This is no secret, but the frequency at which these things are poorly executed may surprise you. Sites with active nursing informatics departments had the best implementations. The farther down the scale in nursing involvement, the worse the implementation. Conclusion: IT (and many vendors) just don't understand clinical workflows.
From the Physician study, it seems vendors are still in love with the physician market - not that there's been a lot of adoption to encourage such interest. The inclination to "follow the money" does not fit this market segment. While physicians account for almost all the revenue generated in a hospital, the vast majority of these docs do not work for the hospitals. In fact, in most community hospitals there is this an unhealthy co-dependent relationship between physicians and hospital administration that works against consistent IT usage.
Few hospitals have any reason to buy technology for physicians, nor to expect physicians to actually use what the hospital may buy for them. These studies show a hopeful trend. Hospitals are increasingly enforcing IT usage among attending and consulting physicians. Some hospitals are even suspending physician privileges for non compliance. Older docs are getting more comfortable with technology, through cell phones, the Internet, and email. The big are of contention uncovered by the study is private practice, and whether the physician wants to change or invest in technology like a practice EMR.
Surprisingly, 75% of physicians use smart phones. Usage though is limited to communications, personal productivity, and the occasional reference tool (if no other reference is available). There's lots of vendor hype around smart phones and similar devices. For most uses the screens are too small, and vendors aren't optimizing their user interface for these kinds of devices.
Here's an except of data from the studies published by iHealthBeat:
Forty percent of nurses surveyed in September said health care IT
vendors were delivering high-quality solutions targeting nursing
workflow, up from 28% in June 2004, according to a survey by Spyglass
Consulting Group.The September survey also found that 64% of
nurses surveyed said they were concerned that their facilities lacked
the appropriate infrastructure to support point-of-care technology,
including reliable networks, interoperability and security
About one-third of nurses surveyed in September
said they were ready to adopt computing applications at point of care.
This represents a 325% increase from June 2004, when just 8% of nurse
surveyed said they were ready. However, just 34% of nurses said their
organizations are investing in point-of-care computing technology, down
from 35% who said the same in June 2004.
The need to improve patient safety and reduce the risk of medical errors drive investments in point of care automation. The market has made progress, but little or nothing has been done to improve clinician workflow. Effective solutions must provide
timely access to the right information to support better and more timely
decisions at the point of care - that means good workflow.
Current deployments of meds administration and nursing documentation apps show increased utilization, but also that they are not being used as intended. The hardware to support these apps are both
fixed location devices (nursing stations, COWs) and mobile devices.
Based on findings from both studies, it is clear that there is no one killer device. The best device depends on the environment, personal preference, tasks to be performed, urgency of the situation, and how well the complexity of the application is matched with the device - you're not going to do patient charting on a Blackberry.
You can read about some other Spyglass studies here and here. Pictured right is the Emano Tech tablet computer - not perfect, but getting close.
It’s interesting that the perception and actuality of wireless LANs differ so drastically. From what I’ve seen, it is true - and unfortunate - that the vast majority of “mobile” solutions in the market are either purely Web-based (better start looking for that taped X) or entirely offline (“now if I can just find a place to synchronize this thing…”).
It’s understandable that many vendors don’t want to soak the costs of optimizing for individual devices/platforms, because in some ways, that’s where the power is. Devices such as the BlackBerry have come a long way, and it’s now possible to do a lot of data-heavy processing over a wireless data plan - which mitigates the need for WiFi and hospital firewall openings nicely, provided the communication is done using a secure protocol such as HTTPS.
It’s true that it’s difficult to capture workflow as complex as patient charting on a little BlackBerry screen. With enough ingenuity, though, we’ve found that it’s ideally suited for simpler (but still mission-critical) functionality like patient hand-off and charge capture.