leonth

The answer may be one of:

* US pharma is developing a me-too / copycat drug that has very similar molecular structure

* US pharma has tried and it failed before public announcement (oh there are so many of them)

* US pharma negotiated contract and found that the Russians are non-negotiable :)

Imagine that there were no FDA Iron Curtain and doctors prescribed afobazole because "the paper was good" (or more likely because the money was good). Then people taking this drug got hospitalized left and right (or - shudder - found dead) because of side effects.

FDA could not do anything to protect her dear US citizens. She couldn't sue the manufacturer or ask them to do further research or impose black box warning or pull the drug out of the market because it was simply beyond their jurisdiction. Doing any of that would constitute a version of the Iron Curtain.

So FDA is doing the right thing to prevent this from even happening by requiring the manufacturer to have everything in place before the drug sees the market.

Furthermore, a randomized controlled trial is no synonym of "drug is good to use". Do they work on US populations? How big is the sample size? Is the methodology sound? What is the side effect profile? Is there any long-term data (like 1 year after using the drug)? These are answered for FDA-approved drugs, but not necessarily for non-approved ones, especially if you are not a healthcare professional that speak Russian (even then, they have teams of professionals looking at drug applications, are you sure you're not overlooking anything?).

Haha. Fancy seeing your post again :)

I'm a pharmacist, but since my boss knows I'm good with computers, she put me to manage the IT systems used by the pharmacy department. This role is well-defined and there is at least one full-time staff in each public hospital here doing this, but no software development knowledge is required because there is a layer of IT consultants that supports us on very technical things (e.g. server provisioning, network configuration, integration). What I do is mainly defining requirements and provide content (both clinical and non-clinical).

I see HN as a way to keep up with IT in general - especially that all my other colleagues talk about totally different kinds of news. I find discussions well-balanced most of the time, and sometimes wonder "what do HN commenters say?" when I get to a news using other routes.

Yes I am a wantrepreneur, if there is such a word. It's pretty cool to (have the perception of) knowing two industries and try to piece them up together.

The article glosses over the fact that wards rounds are typically an exercise that requires vast amount of highly dimensional data. Hidden in the article is the requirement that the information - lab values, vitals, patient location / status, problem list - must be maintained electronically. This by itself is already very hard, especially as paper records never have downtime / network issues and do not talk back when invalid values are written.

Please be assured that the state-of-the-art is not as described. Sufficiently advanced hospitals would have means for healthcare workers to access vitals, lab values, patient location, current medications, current problem list / diagnoses, medical history from previous days and even previous visits, and even medical history from other institutions - all electronically. Staffs enter data into system directly and no transcribing is required. These are available as discrete data i.e. not freetext strings only decipherable by humans.

Also don't forget about electronic prescribing of drugs, glorified vending machines with pockets that only open when there is an order for that particular medication, and barcoded medication administration system.

I don't think OCR will work that well in this case. Assuming the OCR is able to detect the letters, at least from my experience, they like to draw stuff, write things in uneven 2-column or 3-column format, and make use of shorthand symbols.

> the OCR is shown to the google glass wearer who can confirm that the OCR is correct.

This takes a LOT of time. For a page of patient note full of lab values, one needs to make sure each number is translated correctly. And if something is wrong, how do you expect the wearer to fix the OCR results on the google glass?

Price wise, I would think 1 or 2 google glass would be much cheaper than a real EMR system.

Do you see value in having longitudinal / long-term trend of the pulse, BP, etc? Intra-day variance may be very high, but wouldn't there be value of having an average of 100 measurements over a month, as compared to the 2 measurements that you do in the clinic?

Doesn't mean that the doctor needs to see the data as it goes in - you can review the average BP over 1 month when the patient comes for consultation, for example.

Nice to hear back from you. Seriously what are the odds of meeting another pharmacist in HN? :)

Your situation seems to be very fortunate and I'm sure that heading to pharmacy is something your parents are pushing for :) Since you're a partner already, I believe that in the coming years of working in your pharmacy you will be able to apply even more of the programming skills to help operations.

I think if one has a desirable portfolio of projects and good referral, getting a job in IT should still be possible even without a relevant degree. Keep also in mind that not all IT work needs programming on a daily basis unless you are a real hardcore engineer. From what I see there is as much opportunity in vendor support, project management, integration (need technical chops but more on the ops/networking side as compared to programming), at least in enterprise IT here. People already working in these sort of work can very well be less proficient in programming as compared to you right now.

Aside - here not all students do thesis because of the limited number of professors, but I did a shorter thesis-style project about finding gene signatures that may give rise to resistance to imatinib (Glivec). Used a modification of Support Vector Machines (SVM) and recursive feature elimination. It was fun and took the life out from my old laptop because of the heat from crunching the numbers :) Want to hear about your thesis as well :)

Greetings. I was exactly at the same position two years ago but decided to get my license and work as a pharmacist instead.

I am not sure how invested you are against working as a pharmacist, but I would humbly suggest that you at least do that for a while (perhaps 1 to 2 years). Being both an experienced healthcare professional and hacker/developer enables very wide career opportunities. You will be the most tech-savvy pharmacist in the department/company. You will be the most efficient in any kind of data analysis (because your peers are having trouble learning to use Excel Pivot Table while you are writing scripts that take the files as input and perform 1000 Pivot Tables in a second). You will be tasked to be the go-to person for anything remotely related to the million-dollar enterprise software the pharmacy uses. You will get to be involved in the most technologically advanced project the pharmacy is having (think automation, robots, electronic medical records).

In the startup world, the professional status gives you instant, significant positive reputation gap as compared to your usual competent-developer-without-domain-knowledge kind. Not to mention the advantage of having extensive domain knowledge should you manage to find an idea that somehow involves pharmacy.

After that, you can slowly think to get your CS degree (as I am thinking of getting in the near future), or just do the Coursera courses along the way. As you would have some experience working as a pharmacist already, this credential (plus all the experience being involved in pharmacy IT projects) would stay even after you get the CS credentials.

[PS: I'm in Singapore working in a public hospital here.]

> Many of the problems in healthcare involve two systems that need data from each other.

This. From my limited experience, even with all the systems able to talk HL7, apparently what message segments the system sends, the message segments the system reads etc. are different from system to system. The result is entire subsystems or heavy customizations for redirection, massaging, and transformation of messages. Something to think about systems integration.

Another approach is to buy subsystems from as few vendors as possible (naturally subsystems from one vendor would all talk among themselves very nicely) - but that is obviously not without any repercussions.

I'm a pharmacist based in Singapore working in IT side of things for the pharmacy department in a public hospital here (means: pharmacy system, EMR, inpatient automated system, decision support). Still learning but would be happy to keep in touch.

What you describe here has been happening in Singapore public sector for quite some time. Most patients in a hospital will fill their prescriptions in the hospital pharmacy anyway to enjoy subsidies.

Some patients still prefer to have some kind of paper with them, so the doctor will actually print out the prescription but the pharmacy will still process it from the system anyway. When the pharmacy is not busy, we will also pre-process pending (and short) prescriptions from the system and the patient would not need to wait that long for the medicines. It's also a boon for the pharmacy operations because they can smooth out workload throughout the day by pre-processing prescriptions before the peak hours.

Yes it is. It is not about the cost of comparing 2 products, it is about market saturation. When 2010 came, drug A would be off-patent and evidence for using A would be piling up. As a doctor that cares about his patient, one would very reasonably choose drug A over the still-on-patent drug D which is only slightly more effective but since it's new the evidence is weaker (as in -- will this work for the relevant genetic makeup, disease state profile, microbial resistance pattern for antibiotics, is this better efficacy real and not just some shady number crunching by big pharma etc.), nobody exactly knows the long-term side effects, and it's much more expensive. Do you think that the pharma would spend millions to face this bleak market?

(EDIT: this assumes that drug A works "satisfactorily". Drug D still have a chance if it has markedly differentiating features, say if drug A is injected while drug D is taken orally.)

(as a side note, cancer drugs are hot because of precisely opposite of these -- most are fairly new, those that are old are not exactly pleasant to take, there is no "silver bullet" found yet so there are lots of room for improvement.)

Another take on the market saturation is that when the current therapy is simply already good enough, ie. the "silver bullet" has been discovered. That's the reason why there has been no new drugs developed for headache in recent few years (or pain management drugs in general). It's a totally different story when, say, aspirin was first introduced: there is a lot more room to improve on / differentiate from its side effect profile, pharmacological properties, etc.

It does not go down. The more drugs we know, the harder it is to prove that the new drug you propose is better than current therapy. A proposed molecule that does not have significant advantage compared to the current gold standard will never sell. To put it in another way: all the low-hanging fruits are already taken.

Add to the fact that science progresses; we now know more about the human body and generally everything compared to 20 years ago, so more stringent rules and testing are required.

Classic example: before thalidomide, nobody cares about how drugs work in pregnant women. Now all drugs have pregnancy category (how safe it is to be used by pregnant women) approved by FDA, and figuring out this pregnancy category comes with a cost.

I am not terribly familiar with the process, but I assure you it includes: early computational simulations of the prospective molecules; testing on lab animals; testing on healthy people; testing on small (a few hundred) number of patients; large scale (a few thousand) multi-centre trials across the whole world (if you want to sell it to the whole world, that is). This process takes quite a number of years.

Now imagine how many labs, hospitals, scientists, doctors, statisticians, patients (oh they do get compensation), and auxiliary clerical workers need to be paid for the whole process.

Truth is, clinical trials are mind-bogglingly expensive and takes years for turnover. Also some newer drugs (I'm not sure about Nexavar) are rightfully expensive to manufacture due to requirements in equipment, sterile conditions, etc.

If I were to say anything negative about this it would be that this sets a bad precedent for any new drugs coming to India -- knowing that the government can effectively void out drug patents after certain number of years, the big pharmas probably would not want to set up shop in the country altogether. Does that sound familiar?

This town is no longer friendly for business.

I am not a clinician (yet), but the result of a study of 11 people is hardly "actual evidence". Even if it actually does all these things, you never know what bad things will happen when you take resveratol more than 30 days. "Over 400 genes with altered expression" sounds like a lot of side effects to me.