CIO 1-1He was trained a medical doctor, and has taken an oath to heal the sick. Today, Dr. Colin Quek prescribes remedies of a different kind.
Now the CIO of Raffles Medical Group, Dr. Quek cares for the health of IT systems and diagnoses ways to enhance the healthcare group's business processes through the use of technology.
The doctor sat down for a candid interview with ZDNet Asia, where he discussed why the healthcare industry cannot afford to be extravagant with its IT investments. He explained that healthcare carries a unique social component, making it difficult for market players to increase their service charges in order to fund their IT expenditure.
Dr. Quek also noted that there is a limit to how technology can help reduce cost simply because healthcare is a "high-touch" industry, and unlike the manufacturing sector, technology cannot replace human contact.
Q: What would you say is unique about heading a healthcare provider's technology direction compared to a CIO of, say, a retail company? Quek: The general IT needs are common across all industries, issues like finding the skills to manage the types of technology implemented in your organization. On top of that, there are domain specific aspects like healthcare-centric needs and applications. Similarly, banking would have its own special set of applications.
Quite a few in my team didn't come to us with prior healthcare background. When we hire a new person, it's helpful if they have worked in the healthcare industry because they're more attuned to peculiarities of the domain, but it doesn't prevent the person from entering the healthcare.
We can put all the antivirus and anti-Trojan tools, but the probability is high simply because there are that many more Windows servers out there, which makes it a more interesting target for malware, versus a Unix platform.
What's the common gripe among your peers? Not having enough money to do all the work we need to do, and not having enough resources to do all the work we need to do. That's a common bugbear of present day life. Compared to life and user demands 10 to 20 years back, clearly the pressure has grown particularly in the area of cost control and management.
What's the one thing that the tech vendor industry still fails, in terms of addressing the requirements of healthcare service providers? One of the things I've noticed through the years, and which unfortunately hasn't changed a lot, is that tech vendors who dive into the healthcare industry must know exactly what they're getting themselves into. All too often, completely new entrants don't understand enough about the peculiarities of healthcare, and that results in them using the wrong benchmarks and templates when quoting for jobs.
Then when they get the job, they find that they have totally underestimated the amount of effort required and they die trying, and the customer suffers also. So it's a no-win situation, and it hasn't improved all these years I've been in the industry. Everyone thinks there's tons of money to be made in healthcare, but nobody fully understands just how complex a domain it is.
But does that mean you end up going to a player with more experience, but typically then costs more? I guess there's always a price for experience. But I think it's in everybody's interest to succeed rather than to fail accidentally. Competitive pressures will keep vendors honest about pricing. But at the end of day, people are too focused on getting the business at the price the customer wants, and the sales person doesn't talk to the delivery person as to what exactly is involved to deliver the goods. That very often creates a very big problem and gap between what was promised, and what can be done and how much it would take to get it done. This mismatch is still very prevalent today.
Also, relative to other industries, healthcare spending pales in comparison to the financial industry, for example. Reality is, healthcare is unlike other business because there's too much of a social component to it. Much as there may be opportunities to invest big dollars in technology, it's severely moderated by how the cost of those investments can be recovered. You can't just raise prices of healthcare services to get the returns back. There's a balance that needs to be struck, unlike financial industry where they make oodles more money from a turnover and profit margin standpoint. They have better opportunity to reinvest. I don't think we have such big margins in healthcare.
From doctor to techie
Name: Colin Quek
Job title: Chief information officer, Raffles Medical Group
About Colin: A medically-trained doctor who took on his first CIO role in 1997 at Singapore's Tan Tock Seng Hospital. He moved in May 2000 to assume the same role at the National Healthcare Group, before taking on his current position with Raffles Medical Group in early-2006.
About Raffles Medical Group: Founded in 1976, Raffles Medical Group is the largest private medical group in Singapore, operating a network that includes 65 family clinics, a private hospital and insurance services. It serves over 1 million patients and 5,500 corporate clients, and also runs three clinics in Hong Kong.
That's a tough one, balancing social interests with the cost of providing healthcare services. Is there any way technology can help soften that? I guess the promise of technology is always about how they can contribute to cost management because it will improve efficiencies--but it can only do so up to a point. Healthcare is still very much a human-intensive industry. When people are sick, you still value the touch of a fellow human being. It's not something technology can replace all together. This hasn't changed all these years.
The biggest cost to an organization is still people. So you have a high-touch industry where naturally your cost structure is skewed to an expensive resource. You can use technology to keep the business going, but it can never bring about the kind of cost savings you see in manufacturing where the tech replaces the human.
What have you been keeping busy with? We are wrapping up a project that has taken up most of our time over the past year, and we'll be rolling it out before end-2008. We're implementing a new system that will consolidate three of our business systems into one.
Because of the organic growth of the company, we've had to add new systems for each new line of business that was introduced through the years. Today, we have the hospital system, specialist outpatient system and GP (general practitioner) system. When the implementation goes live later this year, there will be one common system to service all three functions and become the common patient-facing application platform for the next five to 10 years.
We're also taking the opportunity to do an infrastructure refresh to ensure we have the means to support the new system, and prepare us for future growth as the organization expands into the region. We have other minor systems that run the medical lab, the appointment booking system and corporate billing requirements.
What will the system consolidation involve? It will run off a pair of Unix servers so we're moving from the Windows environment, which we are currently on, to IBM AIX as the platform for this particular application.
So you're switching operating systems for this? Yes, though Windows still remains the primary platform outside of this application. I guess you can call it a perceived risk that we want to manage. The Windows environment is seen to be still vulnerable to viruses and Trojans. And given that, in the future, all our eggs will be in one basket, we opted to take advantage of what I'll describe as "security obscurity" by deploying a Unix platform versus a Wintel (Windows-Intel) environment.
It's just about managing risk. Today, if one subsystem fails, only one part comes to a halt. In future, when we put all three business functions on one system, and that system goes down, the whole organization comes to a halt. So it's part of our strategy to minimize the risks. We can put all the antivirus and anti-Trojan tools, but the probability is high simply because there are that many more Windows servers out there, which makes it a more interesting target for malware, versus a Unix platform.
We think we're getting better protection by going with Unix, and the Unix track record in uptime is strong. Although the Windows camp would argue the gap is no longer as wide, the concept of a "therapeutic reboot" (rebooting the system every now and then) is still considered to be good practice for Windows.
So the platform switch would require retraining? Yes, there's a lot of reskilling involved. The infrastructure team has gone through at least 10 to 15 man-days of training since September last year. The developers have probably clocked about 20 to 25 man-days of training. So, it's substantial re-training to familiarize them with the new platform. The infrastructure guys have been predominantly a Windows shop all this while. We have a Hewlett-Packard Tru64 Unix server that we hope to decommission soon. With the re-platforming, they will need new skills, for example, to learn how to do systems administration on Unix.
On the application side, again, it requires re-platforming. Much of our in-house applications are developed on FoxPro, but we're moving to Cache because the new application was built on that--though we have been doing a fair bit of development on .Net as well. Cache has been around for a long time and its root goes back to MUMPS. It underpins quite a few healthcare products out in the market, such as Epic Technologies, including its EMR (electronic medical record) product. A few lab systems are also written on top of Cache. Moving forward, our applications will probably be built on a hybrid of Cache technology and .Net.
We'll be moving to a new platform, TrakCare--developed by Trak Health--which will not only service patient management and patient accounting needs, it's also going to introduce a new lab system and radiology information system. We're going to be changing practices and embark on electronic order entry for our clinical staff. And there's the prelude to making our move into EMR in 2009.
I will also be initiating another project for implementing backroom integrated ERP (enterprise resource planning). Currently, we have a separate financial system and other related systems in bits and pieces. Now that we've got Trak supporting all the patient-facing front-end operations, the next piece to look into is integrating the backroom including finance, HR, management and payroll, into one platform. Previously, it's all small piecemeal systems. Now that we've started to tidy up the frontend, we have to do the same in the backend.
It is a massive change for the organization, and one that cuts across all levels. But it's a necessary step to refresh and prepare the organization for our next five to 10 years.
What benefits are you expecting from the new deployment? The payback will come more in the form of better integration of information, particularly about our patients. We will be able to have a single record of a patient, both from a financial and administrative perspective, as well as in the future, from a clinical perspective.
Today, if a patient is seen in one of our GP clinics and for whatever reasons, they don't feel better and visit the hospital at night, the hospital doctor won't have the benefit of seeing the GP's notes because the clinic is already closed and the notes can't be faxed over to the hospital. As we get into EMR in future, this will no longer be the case. The aim is that any doctor seeing the patient will have better visibility of the patient's condition.
In the longer term plan as we extend into the region, technically speaking, there's no reason why a patient cannot receive a similar level of care if they were to visit one of our facilities overseas. So it's this continuity of care that patients can benefit from the new technology.
Administratively, also, the patient will register only once when they see us. Today, because the systems are separate and disparate, they have to undergo re-registration each time they touch a new part of our business. So in the future, we'll have one common record and that directly opens up all sorts of other possibilities. For example, the patient could pay his hospital bill at a GP clinic near his home.
So clinics under the Raffles Medical Group will also be included in the new system implementation? We don't have a choice. Because of the way our business is intertwined, there's no practical way for me to do a piecemeal rollout. There's a lot of interaction between all our healthcare subsidiaries. For example, when a company signs up with us to provide their employees GP, specialist and hospital services, at the end of the month, we have to consolidate the utilization across all three service lines and generate one single bill for the company. If I were to break up the Trak implementation, then for each new part that I roll out, I have to build a way to bridge that information. That creates a lot of unnecessary load that will be thrown away in a month or less.
So we don't have a choice but to go with a big bang approach, because the integration piece makes it next to impossible to decouple one from the other.
What do you think of all the new technology in healthcare like telemedicine? Well, it's always been said that healthcare has been a laggard in the adoption of technology. I think it's not so much that we're adverse to the adoption of technology, but it's always about finding the right reason to adopt and put in place a certain technology.
For us, being a for-profit organization, I think we look very hard to make sure we have a clear payback or need for the technology before we put our money into it. We don't really have the luxury of being on the experimental cutting-edge of things. So we are quite clear that there has to be a clear economic payback for us before we sink our dollars into it.
So things like telemedicine, we'll look at it and see what is the fit. My own personal view is that this little island is too small to invest in a lot of telemedicine. Yes, it has its place in the SARS era, for example, where due to the need for physical isolation of people, technologies were deployed so families could still stay in touch. But today, does Singapore really need technology to allow a patient to see a specialist? It's probably easier to bundle them in a taxi and take a 30-minute ride to that medical facility. Notwithstanding, there's still a value for a select subset of people, people who are bedridden where it's simply not convenient to bundle them in an ambulance just to go see a doctor. It may well be possible for a nurse or doctor to do a check on them over a videoconference link.
But the reality is that, I think telemedicine can only work up to a point because the tech does not cater to the full range of senses that the doctor uses in the medical care process today. Color fidelity is still debatable. Whether a patient is pale looking, or whether an area on the body is red and inflamed--these are critical things for a doctor to notice, but which may not be faithfully reproduced. Nobody talks about medical-grade video conferencing. Similarly, the sensation of touch in the examination process, to tell you whether there's a bump or feels tender to a person. All of these are absent in telemedicine. It can have a role, but I find its application quite limited today.
How about RFID (radio frequency identification)? A lot of potential. It can do a lot particularly, for example, to ensure the right drug is given to the right patient. You can tag a person and a drug, and the nurse can use the data to check the time and means to administer the drug, and minimize the risk of giving the wrong drug to the wrong person at the wrong time. And if an elderly person wanders away in the hospital, you can track her easily instead of sending a batch of nurses and attendants to hunt for her.
What's holding it back? Cost. Even though the cost of RFID tags is down to cents, it's still money. The bigger cost lies in equipping your facility with enough transceivers to make the tracking useful. There also aren't enough common coding standards for drugs and drugs today also don't come pre-tagged from suppliers, so there's a lot of additional work to be done if we want to deploy RFID.