Can a Client Server EHR Provide All the Same Benefits of Cloud EHR?

One of the most popular battles discussions we’ve had on this site since the beginning is around client server EHR software versus cloud EHR software. It’s a really interesting discussion and much like our US political system, most people fall into one camp or the other and like to see the world from whatever ideology their company approaches.

The reality I’ve found is that there are pros and cons to each side. Certainly cloud has won out in most industries, but there are some compelling reasons why cloud hasn’t taken hold in many parts of healthcare.

With that in mind, a client server EHR vendor asked me to list out the reasons why someone should go with a Cloud EHR over client server. Here’s my off the cuff responses:

No IT Support Needed beyond desktop support – This is a big benefit that many like. Plus, they add in the cost of the server, the cost of the local IT person and so they see it as a huge benefit to go with cloud software

Automatic Updated Software – Not always true with the cloud, but they like that the software just updates and they don’t have to go around updating software. Of course, this also has its downsides (ie. when an update happens automatically and breaks something)

Small Upfront Cost – Most Cloud solutions are billed on a monthly charge with little to no upfront cost. We could argue the accounting pieces of this and whether it’s really any better, but it feels better even if many cloud providers require the 1-2 year commitment. In some large organizations this type of payment plan is better for their accounting as well (ie. depreciation of equipment, etc)

More Secure – Obviously this could be argued either way, but those that believe cloud is more secure believe that a cloud provider has more resources and expertise to make their cloud secure vs an in house server where no one might have expertise

More Reliable (backup/disaster recovery) – Similar to the secure argument as far as expertise and ability to provide this reliability

Single Database – There are cool things you can do with data when every doctor is on one database and one standard data structure.

Available Everywhere – At home, office, hospital, etc. (Yes, this can be done by many client server as well, but not usually with the same experience).

I’m sure that a cloud EHR provider could add to my list and I hope they will in the comments. As I was making the list, I wondered to myself if a client server EHR vendor could provide all of the benefits listed above. Let me go through each.

No IT Support Needed beyond desktop support – Some EHR vendors will do all the IT support for the user. Plus, it’s a little bit of a misnomer that you need no IT support with a cloud hosted EHR. You still need someone to service your network and computers. More importantly though, most client server EHR vendors are offering a hosted EHR option which basically provides this same benefit to a practice.

Automatic Updated Software – More and more client server vendors are moving to this approach for updates as well. This is particularly true when they offer a hosted EHR environment where they can easily update the EHR. It’s a different mentality for client server EHR vendors, but it can be done in the client server environment.

Small Upfront Cost – We’ve seen this same offer from almost all of the client server EHR companies. It’s a hard switch for EHR companies to make the change from large up front payments to reoccurring revenue, but I’m seeing it happening all over the industry. The only exception might be the big hospital EHR purchase. In the ambulatory EHR market, I think everyone offers the monthly payment option.

More Secure – This is one that could be argued either way. Either one could be more secure. Client Server vs Cloud EHR doesn’t determine the security. A client server EHR can be just as secure or even more secure than a cloud EHR. I agree that generally speaking, cloud EHR is probably more secure than client server, but that’s speaking very broadly. If you care about security, you can secure a client server EHR as much or more than a cloud EHR.

More Reliable (backup/disaster recovery) – Similar to secure, you can invest in a client server infrastructure that is just as reliable as a cloud EHR. It’s true that a cloud EHR vendor can invest more money in redundant systems usually. However, a client server EHR vendor that hosts the EHR could invest just as much.

Single Database – This is the one major challenge where I think client server has a much harder time than a single database cloud EHR provider. Sure, you can export the data from all of the client server EHR software into a single database in order to do queries across client server EHR installs. A few vendors are doing just that. So, I guess it’s possible, but it’s still not happening very many places and not across all the data yet.

Available Everywhere – This can be done by client server as well, but the experience is often a subset of the in office experience. Although, this is rapidly changing. Bandwidth and technology have gotten so good, that even a client server install can be done pretty much anywhere on any device.

Looking through this list, it makes a great case for why client server EHR software is going to be around for a long time to come. There’s nothing on the list that’s so compelling about cloud hosted EHR software that makes it a clear cut winner.

As I thought about this topic, I tried to understand why cloud’s been the clear cut winner in so many other areas of technology. The answer for me is that in our lives portability has mattered a lot more to us. In healthcare it hasn’t mattered as much. Plus, new client server technologies have been portable enough.

Long story short, I’m a fan of cloud technologies in general, but if I were a provider and a client server technology provided me more features, functions, better workflow, etc, than a cloud EHR, I wouldn’t be afraid to select a client server EHR either.

Also worth clarifying is that this post outlines how a client server EHR can provide all of the same benefits of a cloud EHR. However, just because a client server EHR can provide those benefits, doesn’t mean that they do. Many have chosen not to offer the above solutions. Although, the same goes for cloud EHR as well.

What do you think? Are there other reasons why cloud EHR technology is so much better than client server? Is there something I’ve missed? I look forward to reading your comments.

About the author

John Lynn

John Lynn

John Lynn is the Founder of the, a network of leading Healthcare IT resources. The flagship blog, Healthcare IT Today, contains over 13,000 articles with over half of the articles written by John. These EMR and Healthcare IT related articles have been viewed over 20 million times.

John manages Healthcare IT Central, the leading career Health IT job board. He also organizes the first of its kind conference and community focused on healthcare marketing, Healthcare and IT Marketing Conference, and a healthcare IT conference,, focused on practical healthcare IT innovation. John is an advisor to multiple healthcare IT companies. John is highly involved in social media, and in addition to his blogs can be found on Twitter: @techguy.


  • Thanks for sharing your views about cloud EHR.Another point that I would like to mention is ease in integrating cloud EHR systems with mobile taking EHR system anytime anywhere for patients and physicians.

  • Hi Satin,
    Yes, I covered that a little bit in the “Available Everywhere” section. There are ways around this on the client side, but they are harder to scale. I’ll be interested if it becomes so hard over time that it does become an issue.

  • Vendors, us included are increasingly pushing cloud. ,it benefits us from a financial and support standpoint. However the jury for me is out on end users, see most providers received good response times on in office systems and if Internet was down they were not, even large practices with private lines.

    Cloud lowers overall TCO for IT for smaller to medium sized practices, but often at a cost to performance. Newer speeds in communication have greatly eased this issue as has technology for WAN acceleration and APP centric client rich cloud solutions.

    However any gain to the practice in TCO is usually out weighed by the cost for SAAS/ASP cloud hosting software license fees by the vendor. That being said some vendors, us included, are leaving the decision to the end user or offering hybrid solutions but forcing the additional license fees to be monthly regardless of end user decision on Application Delivery.

    To me the cost of Healthcare software and support is rising not decreasing so the monthly model and increased costs do need to be transferred to the end user regardless of the IT infrastructure.

  • On the deployment side cloud is the clear cut winner. I saw that with a 18 doctor GI group we signed three weeks before October of last year and was up up an running by Oct 1 and met MU stage 1. Installation was done in a week, practice was up and trained in 3 weeks. Never happen client server, they had over 200 computers and 8 locations, 3 asc.

    Of coarse the software is also key here, if it is not real easy to use no way a large group gets up and running so fast.

  • Yes; a Client Server Infrastructure can do almost all the things that a SaaS model on cloud can do. Yes; you can have the CITRIX Licenses and remote into the app from any where. Everything can be done. Having said that, everything costs $$ including CITRIX or similar licenses; and having to host the PHR separately – which also can be done at a cost.

    Although the Vendor can host the servers and do all the maintenance, there is a cost associated with that and a cost associated with the refresh of such servers every 4 years or so.

    Can the Client Servers do it? Absolutely yes. Its like fitting a square peg in a round hole.

  • In general, one of the biggest factors favoring the move to cloud was/is performance. For many, cloud-based, DB intensive systems are “fast enough”.

  • Another factor favoring cloud post-deployment is the ability to spin-up and spin-down computing resources. Relatively easy and quick to add or subtract disk space and memory for example.

  • I have no personal experience with this, is there a significant difference in module load times or screen updates between cloud and client house. Seems to me there would be a quicker time with client server. This could make a significant difference in productivity with wait times, even though each one is very small they do add up to a significant degradation in employee productivity.

  • The discussion is a bit like comparing paper to client server. Sure you can make a case where the benefits are similar but in the end for small practice systems, companies are investing in the future, which is the cloud. Seems like the only people really interested in having the conversation are those invested in client-server and are a bit resistant to change. I would also mention the cost and configuration fees for many of the lab, erx, PHR, erad and HIE integrations are not shared with client server as in cloud and this is a significant factor as we move forward.

  • Great points everyone.

    It depends on how you’ve set it up locally and how you’ve set up the cloud. If you have enough bandwidth and your cloud EHR has done a great job setting up their system, the difference is negligible.

    I’m not invested in client-server and I’m having the conversation. Interoperability is a good one that I should look at adding to the list. Although, this is possible with client server. Plus, it’s amazing how even a cloud EHR vendor has to do unique things for each provider for their interfaces with labs, PHR, etc.

  • I’ve spent about 3 decades in Financial IT, mostly in banks. Long term, I’ve observed that except for very massive systems needing low latency response concentrated in relatively few main locations – ex trading systems, that banks eventually found that they didn’t want to put all their effort into IT rather then into banking. And even in trading systems, the actual systems are now typically co-located in financial exchange computer centers to reduce latency.

    Doing things in house requires piles of internal expertise on issues ranging from networking to database administration. Clouds still need local security and networking people, but the CIO doesn’t have put huge resources into software development, interconnections with other providers, and many other things not central to healthcare.

    However, far too many managers think that the ‘cloud’ relieves them of responsibility for DR, backup, security and far more. It certainly can simplify many issues, but in the end the CIO is still responsible, but he knows, or hopes, that the people doing most of the work maintaining and updating major systems are dedicated to just that and have excellent resources for doing so. And it sure ought to be easier for the cloud vendor to make the connections to other institutions. It should all be standardized and simple.

    Cloud is no panacea, but if properly handled it ought to be a huge improvement over typical client server in most situations.

  • Troy, I believe you are correct, I don’t want to even get into my years in Financial. The problem is that the Financial Industry Software is Licensed to be hosted on a big backend with appropriate per user license fees and runs on a very basic Terminal UI. EHR is not there yet, most the applications are web centric and yet provide a rich UI.

    That being said, I don’t push end user to any of the technology, and yes Citrix/MS Remote Desktop has done a lot to help centralize this, however, the issue is more in the technology, it does work, but has issues.

    I currently have many end users on three technologies, 1.) Client Server 2.) Cloud 3. Hybrid – Power User/Doctor – App Centric (Local Application, Rich Client) running to Cloud Hosted Servers; majority of end users on Thin Client (Remote into Cloud).

    I also worked with a large EHR deployed via Web, one client 180 Doctor 30+ locations, multi-tennenat multi-state, and can tell you many issues based on Java and very high Development Cost for Application Development. End User experience not so good, but easy to manage and deploy as everything in the session was handled host side.

    The industry is going Cloud, end users are in agreement, so that is where we will be. From a technology point of view, the devices are getting very powerful, therefore, the vendors will turn to Application Software (Apps) that will store the data locally in memory and then push the data to a Cloud End Point that processes and stores the session data sent from the Client. The CONCEPT OF WEB SERVER BASED HOSTED APPLICATIONS WILL FALL BY THE WAYSIDE, WHERE ALL PROCESSING IS COMPLETED AT THE HOST. When a Doctor sees 40 patients a day, after going to the Cath Lab/Hospital/ASC and every second is critical, this technology can slow him down, every second an application has latency is a high cost to the practice and frustration to the end user.

    This is when end users will see the performance they get from a localized environment. Many developers are already working on this as it is the clear choice moving forward for Graphics Rich UI Applications to run in a CLOUD environment.

    All you have to do is look at the current devices and applications outside of EHR to see this is the future.

  • John,

    It depends on how you’ve set it up locally and how you’ve set up the cloud. If you have enough bandwidth and your cloud EHR has done a great job setting up their system, the difference is negligible.”

    Bandwidth less a issue then ever before, and dependent on Cost the servers in the cloud are on very high speed low latency lines, with WAN Acceleration technology.

    I 100% agree with your point on the IT Infrastructure standpoint.

  • The cloud makes good sense for smaller practices. The larger you get, the less sense is can make.

    Latency and the disbelief that a backup broadband pipe is needed are two big factors to issues. Until going to a cloud based EHR, many practices had no clue how often they had broadband outages…if you want to see an entire office come to a grinding halt and have them all stare at each other, don’t pay your broadband bill one month.

    Still bigger is the belief that being on the cloud handles all HIPAA/MU related issues…which of course is not close to being true.

  • John,
    Not my experience, For example large group MPLS with Dedicated ATT Internet and Backup COmcast. 0 downtime, high speed RDP sessions, 200+ computers. I think that is more about what John stated above, what technology you have invested in to make it all happen.

    This particular practice has 1 IT person supporting 3 ASC’s and 8 locations servicing over 500 Encounters per day.

    Agreed 100% on the lack of proper Link Technology or methodology to the Cloud, or end point solution (The Hosting Company itself) can be a huge problem for a practice. Like you stated, one serious downtime could literally cost a practice hours on end, plus much frustration rescheduling patinets, etc…

  • Am I wrong, or does all the bandwidth required to handle 200 users simultaneously become cost prohibitive? As far as a reliability factor, much depends on what part of the country in which you reside.

  • It’s occurred to me that there is one potential advantage to C/S. Assume that you just don’t have ‘users’ such as doctors, nurses, techs, etc. You should also have a large number of devices like infusion pumps, ekg’s, pulse oximeters and the like, not just monitoring or pumping but also using ‘intelligence’ to provide appropriate alarms (not just a loud tone), automatic adjustment of IV rates and the like, and while most of these will have fairly low data rates, it all adds up, plus you need quick response to various inputs. This is doable in a cloud setup if you have a low latency high bandwidth connection with a fast responding server – but the farther away geographically the server is, the worse the response will be. Example – you give med X where a patient tends to have a big change in pulse and BP if given faster then the patient can tolerate, so this monitoring allows for the infusion rate to change based on changes in pulse and BP.

  • @Bill,
    Costs for this end user was negligible. They already had fiber for communication and just upgraded to 20mbps for the dedicated Internet up and down. That being said yes they are in South Florida, a large populated region with fiber available. This is not the case for our rural clinic in the western Georgia mountains who had enough trouble getting Internet, fiber and reliability is a long way off, so location does matter a lot in Client Server vs clou decision.

    @R Troy
    You are correct medical devices and monitoring will perform better in a Client Server environment then a computer on a network far away. But generally you would run this local and send data to the EHR.

  • Brendon, makes sense. A huge obstacle in some mainly rural areas is very little bandwidth available for Internet and other services. There are Fed programs that are helping, but a hospital in such an area has a huge disadvantage whether cloud or C/S, but worse for cloud.


  • For every great cloud implementation, I can show a handful of bad cloud implementations.
    The bad is not necessarily based on the fact that the EHR is cloud based, but poor choices being made (works the same with C/S setups).

    It still comes down to money.
    “I don’t want to pay for extra bandwidth”
    “I don’t want to pay for a backup broadband connection”
    “Why is the EHR running so slow?”

    I still believe this is an industry cycle, remember:
    mainframe >> desktop C/S >> thin client (mainframe) >> desktop C/S >> cloud (mainframe) >> anyone…anyone?

  • John,

    Very true. Cheapness by those paying for bandwidth – both LAN and to the web, can ruin any implementation by cloud. And whether you have direct lines to the vendor (which is far more secure then by Internet) or comms are Internet based, you still need 2 or more connections with diverse routing, which many firms are unwilling to pay for.

    Regardless of whether you go for C/S or cloud, quality of service from the vendor is critical. I’ve been dealing with a vendor in the Financial Market Data world for about 15 years, eventually going from C/S to cloud, and I know that vendor support is critical – and sometimes lacking in this immensely complicated product. And I’m just referring to the tech side. But there is also the issue of how end users are supported and trained. Plus who is responsible for software and database upgrades and the like.

    And as someone above pointed out, you may need some sort of hybrid model, where local servers consolidate traffic (such as for medical devices and monitors) but where the overall EHR is mainly cloud based.


  • @ Ron
    For many of the “cloud” implementations I’ve seen, the ease that one perceives from the cloud isn’t there.
    Sure there is no server to maintain, but the client EHR interface is generally either a highly tweaked IE the freaks if a new tool bar is accidentally added (thanks Adobe) OR full blown software that can’t have updates pushed to it, therefore manual updating of every client is required when updates occur.

    So, assuming the EHR vendor isn’t trying to push you to their higher profit margin cloud product, one should expect a solid 5 years from a server.

    We could go on a linux vs windows based server argument here, but no need to bore everyone.

    So, what does the cloud actually provide a typical private practice?
    Offsite backup
    An easier contingency operation option
    Lower IT maintenance fees
    …let’s see, what else…

  • John,

    A cloud vendor SHOULD relieve you of most system and server maintenance, of having a 24×7 IT staff just to handle all the routine parts of data center support, plus putting a lot of effort into supporting something – computers, that is not a part of your base expertise. Plus the up front investment and ongoing maintenance and repair issues, and having your DR site also having to maintain a full system. Hospitals often put data centers into basements; Superstorm Sandy proved how bad an idea that was. With cloud, you need a much smaller space on site for telecom gear only.


Click here to post a comment