Tags: , , , , , , , | Categories: The Science of Think by Chris on 10/28/2013 8:46 PM | Comments (3)

 

  HealthCare.gov is sick. As a thirty year veteran of the information technology business I’ve spent the last fifteen years of my career consulting with fortune 500 companies. My background is data architecture and systems software design. As such I’ve been directly involved in designing and building world class websites just like HealthCare.gov.

 

  As I’ve listened to the debate since the failed launch of the federal website that was supposed to serve as the portal for millions of Americans to shop for “affordable” healthcare it became increasingly clear to me what had likely happened to the ailing website. I think it succumbed to a series of obvious blunders made all too often by large corporations in a hurry to finish a software project on time, on budget or both.

 

  In the crunch to avoid accountability to their superiors, middle management will often wrestle control from the qualified technologists, architects and engineers (in other words the professionals) who are actually building the technology and begin making their own decisions based upon timelines, budgets, perceptions and politics. These are administrators with little or no technical background pulling rank and steering the project in a different direction for political reasons usually which includes making themselves look good or in some cases to not look bad. I’ve seen it many times.

 

  The point is I believe HealthCare.gov is much worse than is generally known and the Obama administration is in major fire drill mode while attempting to contain an otherwise out-of-control situation. Allow me to illustrate and hang on because this will be just a bit technical:

 

  HealthCare.gov does not run on a single webserver. That would be ridiculous and would never carry the traffic load. It actually runs in a web farm containing many web servers all bridged together by a load balancer of some kind. As web visitors arrive at the site the load balancer moves them to a webserver that is not overloaded at the moment. It’s spreading the load around. That is of course if the balancer is working properly. These servers are all racked in a data center somewhere to provide power, cooling and battery backup in case of power outages etc. Web servers are subject to memory spiking demands and consumption due to high volume, spiking cpu and disk I/O loading etc. Presumably the hardware engineers sized the web farm based upon requirements given to them by their customer, Health and Human Services, which is the Federal Government. Obviously the webservers cannot respond to traffic. This can be for a variety of reasons contained within the web farm some of which I’ve already described. The problem is there are several more “layers” of technology that make the whole website work that are much more nefarious.

 

  Oversimplified the internet runs on wires, routers and databases. The webservers execute the code that displays what we see when we surf as do the database servers underneath.  My concern is the underlying data stored in databases which is not being discussed, at least not in the media. The internet is about data and these data are stored in databases, period. Remember my explanation above about “Data Architecture”? Database designers like me refer to it as Data Modeling. We take processes, like selling health insurance to people, and design a data model that looks just like the process. Without it the website would never work. I’m deeply concerned that model may not reflect the process of effectively presenting a product to the American people in the way they’ve been promised. Why? Probably because of poor requirements that came from the customer (i.e. the Federal Government). The database is the foundation of the house. If it is wrong the entire house will be wrong. How can you change the foundation of a house without moving the walls, plumbing, wiring and roof also and avoid tearing the whole house down in the process?

 

  Customers always ask for full featured software, delivered in a few weeks with no problems. In other words, feature rich, bug free and delivered quickly. I usually respond and tell them they have to pick two. It’s rare you actually get all three. I’ve had competitors claim they can deliver all three but in most cases they’re just posturing for business and being less than completely honest, sometimes even with themselves. The website builders and Sebelius got in a hurry to meet the deadline and failure to launch is the result.

 

  Now for the really bad news. Private sector health insurance companies are the masters at statistical analysis. I should know, I work in Nashville for healthcare companies. I design the data models they use and queries that do the analysis. To price health insurance coverage for an individual there are a plethora of “data points” insurers capture to properly identify their risk so they can charge and profit from their business. My understanding is that Obamacare is tied into many other data sources like the IRS and who knows what else in an attempt to track, monitor, report and manage the entire enterprise healthcare effort. This is an enormous data modeling challenge connecting heterogeneous data sources from disparate government agencies that depend on other systems availability, data conversion, data mapping etc. The possibility of copious data errors in the near term is quite high and we haven’t even talked about data and internet security yet.

 

Gold, Silver and Bronze health insurance plans sounds suspicously like "Small, Medium and Large". Really? So one size fits all? Or in this case 3 sizes fits all 300 million Americans? If the data analysis required to properly configure health insurance coverage for an 80 year old male is not properly data modeled and analyzed to suit his needs he could very well end up paying for maternity coverage. We don't know for sure because the website is not working and they're hiding the numbers. The private sector has this down cold (remember profit incentive? Hello?) Why is the government who has zero experience doing this taking it over? Seriously?

 

If the Chinese have been probing and hacking private corporate networks in the last several years (and they have successfully) then HealthCare.gov is a Christmas turkey that has just been served up on a plate and Obama is ringing the dinner bell.

In my early morning Agile stand up meetings with my teams I often tell them, “The same level of thinking that created a problem is insufficient to solve it.” “If you don’t have time to build it right when will you have time to do it over?”

I also tell them “Less is More” when it comes to software. Object Oriented Programming (OOP) is a design principle to REDUCE the amount of code in software. It lowers cost of ownership and reduces bugs. All software has bugs, period.

 

  “When everybody owns it, nobody owns it” ~ c.carter

   

  Government has no incentive to succeed financially. If they need more money they pick the pockets of tax payers. The federal government does not run profitable business. Just look at USPS, IRS, Social Security, Amtrak and now Health Care. Without this fundamental incentive nobody will ultimately take responsibility because after all, “Everybody owns it”. Just look at the finger pointing going on right now around the website failure. Nobody really owns it but it cost the taxpayers millions and millions of dollars and it doesn’t work. I fear we are in deep stuff for the reasons listed above.

 

  As I am writing this right now the government has brought in PRIVATE sector computer scientists to try and fix HealthCare.gov because the government project was a failure. And the government plans on running our health care? My guess is they are putting spit, tape and glue on the website right now to try and get it running by November. They might succeed in eliminating some of the crashes and performance problems. After that data errors and Turkey dinner lie ahead. In my opinion we could be looking at an entire year before this is straightened out if at all.

 UPDATE:

 The below report filed by Fox News on 11/19/2013 weeks after my article posted here:
Fox News Report - HealthCare.gov already compromised security expert says


Update:

This report filed by CBS News on 12/1/2013 over a month after my article posted here. Warns of "bad data" on the backend databases.
Democrats brace for more bad Obamacare news

 

Comments (3) -

Ben Lambert United States on 10/31/2013 8:40 AM Good explanation on the issue. To say the least I'm very worried about this whole thing.
chris United States on 10/31/2013 9:09 AM Indeed Ben. Knowing how internationally visible this whole affair is I think it is just a matter of time before a major data theft occurs on HealthCare.gov. If it goes the way it has been going if/when a theft does happen the Obama administration will likely cover it up and if/when caught lie their way out of it.
Wade Edward Burt United States on 11/2/2013 8:05 AM Excellent analysis of what is likely to be happening. It is clearly the "duck on the water" mentality around this project.

I have also seen the impact of a reactive management style coming from the mid level and above. I agree with you. Working mostly to perceptions, many of them wrong, mid levels (in collusion with executive levels) feel they can "drive" the project to a successful completion. In my view, budget is tertiary to achieving the goal and the technical requirements.

This behavior reduces the technology project to a political tool. Which is what Obamacare's website appears to have been. I like the analogy of "attaching the wings while the airplane is already at fifty thousand feet". There is no good outcome - thus, the failure of the "front end" of the liberal social medicine experiment known as Obamacare.

Throwing money at it won't make it better. The bigger question is about the rest of the program. Has The Affordable Care Act been designed the same way? It appears likely. Using "if you build it, they will come" as the initial guideline, the political elites said "trust us". And many did. But what has come out so far has been a largely dysfunctional set of rules interpreted as guidelines and goals interpreted as suggestions.

The costs to everyone are higher than expected. Availability of insurance has  been shifted. Somehow, the discussion is around cancellations rather than the goal of the cancellations - which is to get people re-enrolled in more expensive plans with less coverage, higher deductibles and fewer providers. Obama promised you could keep your insurance and your provider. Neither one of those promises has been kept.

The dysfunctional website is an indicator of dysfunction elsewhere in Obamacare. With that in mind, is the website dysfunction a red herring? Has the website been laid on the altar to draw attention from all the other issues with Obamacare? Who knows - one thing is certain - leadership failed to deploy a solid solution on the front end. Do you think the back end looks any better? Look underwater. The duck is likely to be paddling very hard........
Comments are closed