The InfoStride Forum

TECHNOLOGY => Computing and Internet => Topic started by: ReadWrite on Oct 21, 2013, 09:31 PM

Title: Healthcare.gov Launch May Be Too Big To Succeed
Post by: ReadWrite on Oct 21, 2013, 09:31 PM
What do you get when you design a website by committee? Something that looks a lot like Healthcare.gov (https://www.healthcare.gov/), the home page for the federal online insurance exchange. The problems plaguing the site are not only a political black eye or godsend for whichever party is observing the fiasco, but also very much a technical lesson in how not to build such a large complex system.

Today President Obama is scheduled to hold a Rose Garden briefing that will highlight the failures to date and outline what administration officials plan to do about the problem. But as more problems are revealed, the possibilities of a quick fix are virtually impossible.

Bandages Vs. Surgery That hasn't stopped some immediate-care efforts from being made. On Sunday, administrators of Healthcare.gov, in an attempt to alleviate some of the problems, blogged that content regarding different insurance plans would be front loaded to a part of the site that comes before the account creation that is key to the eligibility and sign-up process. The site is also now pointing people to email and telephone sign-up contact information as well as to downloadable sign-up forms, to get more people to bypass the beleaguered site and sign up for their healthcare plans in other ways.

These efforts will ease some of the pressure, but without major overhauls of the system, the continued systemic issues will blog many interested people from signing up for health insurance plans.

Opponents of the Affordable Care Act see the website's failings as demonstrable proof that the law itself is doomed to failure. Sen. Marco Rubio (R-Fla.), appearing on Fox News Sunday this weekend, seemed to sum up the sense of frustration about the Healthcare.gov site.

"n the 21st century, setting up a website where people can go on and buy something is not that complicated," Rubio said.

With all due respect to the senator, that's not exactly true. While setting up a simple e-commerce site is a relatively simple matter of plugging in the right tools to a content management system like Drupal or Joomla, there is a big difference in scale between a small online storefront and a gargantuan presence like Amazon. Just ask @WalmartLabs, which is still trying to catch up with Amazon (http://www.nytimes.com/2013/10/20/technology/to-catch-up-walmart-moves-to-amazon-turf.html).

Rubio's comparison is also off because Amazon and other major online retailers have systems that were built over time, organically, not rushed into existence. Nor are most e-commerce sites required to abide by sometimes byzantine legal statues and insurance regulations that can vary significantly be state.

Still, the fact remains that the Healthcare.gov site is indeed broken. And no quick fix is going to be able to save it.

Ignoring The Symptoms From the start, the implementation of the Healthcare.gov site seemed destined for trouble.

No less than 55 contractors were assigned the monumental task of intaking potential health insurance customers with rigorous identity measures, determine their eligibility status for health insurance and any federal and state insurance premium subsidies they might get, enable them to select the plans for which they are eligible, and then allow customers to sign up for the plan they want.

All of this, by the way, under the auspices of regulations that vary sharply by state in a political atmosphere that makes juggling nitroglycerin jars look easy. Overseeing this technical circus is the Centers for Medicare and Medicaid Services, a Federal agency that is reportedly way outside its comfort zone for managing such a large project.

The unspoken part of this website launch is that very rarely does any such complex system get launched successfully as a whole, functioning system. Twitter, Facebook, Google... all very high-traffic and high-demand sites that, while regarded as stable now (despite the still-occasional glitches), were much simpler in features and complexity when they were first launched.

Some estimates opine that over 5 million lines of code will have to be changed to fix the site. Some software engineers might see that number and wonder why their were 5 million lines of code in the first place.

Healthcare.gov is regarded by proponents as too big to fail, but the truth is, it may have been too big to succeed, particularly on launch day.

The Prognosis Time heals all wounds, it is said, and given the abundance of technical expertise in the world, it seems unlikely that these glitches will never be fixed.

But a lot of time is something the Obama administration does not have. Even putting aside the sharpening of knives from the political opposition to the Affordable Care Act, every delayed applicant puts the law's implementation in danger. If not enough applicants sign up, the financial sustainability of the healthcare act is at risk, because premiums will be too high.

Some have argued that it would be best to open source the entirety of the site's code and let the world's programmers have a hand in fixing the problem. While open source is normally a positive method of development, opening the site's code not at this juncture would probably be a mistake. Code that was designed under a proprietary software license is often not well-documented and hard to decipher for open source developers, and that's when the code is well put-together.

Opening the Healthcare.gov code in its entirety would require a learning curve for outside developers that would likely delay fixes for longer than it would take for developers within the project already to just fix it themselves.

The first real step towards a fix will most likely be the appointment of a lead team that will be able to survey the problems and put an action plan in place. Whether this lead will be a company already in the process or some sort of small supercommittee of techies remains to be seen, but it is clear that the whole "by committee" method of design has not succeeded.

Source: Healthcare.gov Launch May Be Too Big To Succeed (http://readwrite.com/2013/10/21/healthcaregov-launch-may-be-too-big-to-succeed)