On Developing for Android

Marco Arment, the host of the podcast Build & Analyze, had this to say regarding converting his popular iPhone app to an Android app:

By the way, it's not like I hate Android.  I just don't think there's much business for me there. And again, I don't think there's zero business, I just don't think there's enough to make it worth the investment I have to put into developing and maintaining that . . .  I just look at the realities of the market, and the realities of the market - and my time - are that I barely have enough time to manage the apps on the platform I support now . . . I have to focus on the platforms I have time to focus on.  I can't add a whole bunch of them.

This pretty much mirrored my own thoughts on the subject.  From the day my iPhone app was approved in iTunes I've received requests to make it available on other platforms.  First it was BlackBerry and Palm, but the overwhelming request nowadays is for Android.  And, I fear the next Windows Phone 8 will be the next big thing.

I feel bad that I don't have an Android app. I really do.  I know it's become a very popular platform.  I'm not neglecting it because I'm playing sides.  I haven't made it because I'm already stretched so thin.

Marco suggested it would take him months to create an Android app, all while ignoring the upkeep on his Web site and iPhone apps.  Marco is a full time developer.  I, on the other hand, am a full time optometrist.  I think it would take me at least a year to create an Android app, probably more.  Why do I think that?  Well, the iPhone app took about 18 months from the day I bought a book on iPhone programming to the day I "shipped" my product.  Although I'm really proud of it, it was by far the toughest programming task I've undertaken.  

Since then I've had a couple children, still work full time (well, I've cut back a little), and still have a website and an iPhone app that I work on nearly every night. I just don't have any time left.

The other consideration is the financial aspects.  As I said, Marco Arment makes his living as a developer.  My income comes from looking at eyeballs.  I've managed to make EyeDock profitable enough to justify to my wife why, when my children get down to sleep in the evenings, I plug away at my computer instead of watching Dancing with the Stars like everyone else.  

However, the biggest reason I design websites and program is because I enjoy it.  I enjoy creating things, I enjoy making something that's useful, and, I'll admit it, I enjoy the positive feedback I get.  I have no idea how many people subscribed to my site because of my iPhone app.  I'm sure some did, but I've always thought of the app as a supplement to my EyeDock.  Perhaps I'm way off base on that - if I created an Android app maybe I'd double my subscribership and it would be well worth it.  That's an unknowable thing, unfortunately.

Even if I could justify it financially, there's still the issue of time.  I've often thought about farming out the development of an Android app to a professional.  However, my experiences with having outsiders do my coding has been less than desirable.  I feel like I've made good things because I'm an optometrist and I know exactly how I need things to work when I'm seeing patients.  An outsider just doesn't understand at the same level.  I can't justify investing my limited resources in something that I'm not confident is going to be made to my satisfaction.

I sincerely do apologize for my lack of an Android app.  I have nothing against the platform.  It's just that the iPhone came first, and I can currently only support one.  This will change if EyeDock grows to the point that I can dedicate more time and resources to it.  Believe me, nobody wants that more than I do.

Why is Every Guy Named Todd a Total Dork?

Why is it that, every time I see someone named Todd on TV or in a movie, they're a total dork?  

Then again, maybe it's not just a Hollywood thing.  After I got to know one of my roomates in grad school he confessed that, before he'd met me, he'd dreaded having a roomate named Todd because every Todd he'd ever known had been a total tool.  Now that I think about it, I don't recall him ever saying that he'd changed his opinion.  

I happen to think Todds are pretty awesome, but TV and movie writers seem to disagree. According to them, Todds seem to fall into one of three categories:

The Yuppie.  Usually a person that's a little wealthy, a little snobby, and usually pretty clueless.  Movies like characters like this because they make a good bad guy, or at the very least, a guy that they can make fun of - and the audience won't feel sorry for them.  

Prime example: Todd Chester, from Christmas Vacation:

The Nutjob. A guy that's obnoxious, a sexual deviant, or just plain off his rocker.

Prime example: "The Todd", from Scrubs:

The Total Freak - a misfit in every way.

Prime Example: Todd Cleary, from The Wedding Crashers.

 

Off the top of my head, I can only think of one movie Todd or, in this case, Tod, that I was proud to share my name with:

Tod, from the Fox & the Hound:

Merry Christmas!

Daddy on a Bobcat

Drew, my 3.5 year old, announced that

"when I was born, daddy was on a bobcat"

While this isn't necessarily true I got a little chuckle imagining that, if it were true, it would look something like this . . .

Ophthalmic Migraines and Cannonballs

At least a couple times a month I have a patient describe a visual episode that goes something like this:

"I was working on my computer this morning when I startedFortification Scotoma - From allaboutvision.com getting these patchy little blind spots in my vision.  The blind spots started more central and had these flickering little zig-zag borders around them.  Over time they expanded and finally left my vision after about 20 minutes.  I didn't get a headache."

Of course the symptoms can vary. Sometimes these individuals are experiecing this for the first time and are understandably very concerned, and in other cases they might just offhandedly mention that this happens a few times a year. 

In either case, once we rule out other possibilities (such as retinal detachments or vascular problems), we will diagnose this as an ocular or ophthalmic migraine.

Illustration of scotoma from Sir W.S. Duke-ElderAllaboutvision.com has a great summary of ophthalmic migraines and I recommend you read their article to learn more about this strange phenomenon.  

What I want to talk about here is these unusual zig-zag disturbances that can be seen when experiencing this type of migraine.  While in optometry school I learned that these disturbances are called fortification scotomas, where "scotoma" (Greek for "darkness") means blind spot.  I never gave much thought to the "fortification" part until a few years later when I was reading a textbook that stated "fortification" refers to the appearance of 16th century French forts.  

This intrigued me.

It seems that medieval fortresses were usually rounded structures that were place on high hilltops.  This worked well as the soldiers defending the fortess could loose their arrows and watch them sail towards their distant invaders.  From the perspective of an invader, well, it was really difficult to attack the strongholds with arrows raining down on you.

Then, the cannonballs came.

By the 17th century technology had evolved to the point where manueverable cannons and gunpowder became the invader's weapons of choice.  When this started to happen, the old fortresses with their rounded stone walls didn't fare so well.

In response, military engineers began to rethink the way fortifications were designed.  In particular, a Frenchman named Sébastien le Prestre de Vauban designed forts that were better suited for the new modes of warfare.  Vauban started his career directing sieges in Louis XIV's army and, after attacking a lot of fortifications, learned a thing or two about what makes a fort vulnerable.

From theotherside.co.uk:

"Knowing how he would attack a stronghold, his solution was to build "star-shaped" forts with straight-sided moats, lined with walls built of local materials (mainly brick in the north). He left no “blind spots” where an attacker could hide. His defenders could fire on the enemy with cannon mounted behind thick walls on the ramparts; and rake the moats with handguns fired through slits in the walls."

Ultimately, French forts ended up looking like this:

As a consequence, I can see how someone with a knowledge of French military history, while in the throes of an ophthalmic migraine, might have compared their visual symptoms to an arial view of a Vauban fort.   Hence, the "fortification scotoma".  I personally would of called it a "sparkly jaggy thingy", so I'm glad someone else beat me to the naming rights.

So, next time you're suffering from an ophthalmic migraine, be reassured: It might be a little disturbing, but at least you won't be susceptible to cannonball-toting invaders trying to lay seige on your central vison.

(you didn't think you were going to get through one of my blog posts without hearing something really corny, did you?)

Ready Player One

I thought I'd do my a book report - after all, I've had nearly 20 years to read for pleasure without draconian literature professors forcing me to write about it!

I'd been hearing about this book in some of the various podcasts I listen to, and, always in search of a good read, I checked it out.

To save you from reading the jacket cover (or the Amazon page - it's the internet age, after all), I'll give you the rundown: The year is 2044, and the world is in the midst of an energy and financial crisis (a little too close for comfort?)  Everybody escapes real life by immersing themselves in a virtual reality world called Oasis.  

Oasis was created by a man who grew up in the 1980's, and, on his deathbed, announced that he had hidden a series of puzzles inside his creation. The person who solves the puzzle will be rich and will win the right to control Oasis.  The protagonist is a teenager who takes the lead in solving the puzzles.  Of course there's an evil corporation trying to win the prize too, as well as an attractive albeit mysterious love interest.  Actually, everybody has a little mysteriousness going on because they all know each other as avatars within their virtual world.

The intersting part is the Oasis creator's clues revolve around the culture of his youth, especially the geek culture of that era.  As a consequence the youth of 2044 become obsessed with the latter half of the twentieth century and spend inordinate amounts of time "studying" Pac Man, Schoolhouse Rock, and John Hughe's movies.

Some parts of this book really struck a chord for me.  The inventor of the Oasis was born in 1972 (as was I), so it seemed like it was tailor-made for a 39 year old nerd who came of age in the 1980's.  He mentions receiving a Atari 2600 in 1979 (as did I).  There are even references to storing data on analog tapes, Dodge Omnis, and paying 93¢ a gallon for gas (freaky coincidences, or has the author been stalking me?).

Was it a good book?  Well, to be honest, the writing of the book reminds me of the style I see in fiction aimed at juniors.  Perhaps it's because the main character is a teenager, but it seemed a little strange because the book is obviously going to be most enjoyed by people in their late 30's.

However, it's hard not to enjoy all the references to things I loved in the 80's , some of which I haven't thought about in quite a while:

TV Shows: Family Ties, The Greatest American Hero, Airwolf, A Team, The Greatest American Hero, Misfits of Science (does anyone besides me remember that one!?), Buck Rogers, Silver Spoons

Video Games: Pitfall, Zaxxon, Galaga, Q Bert

Movies: War Games, Real Genius, Better Off Dead, Evil Dead, Vision Quest, Explorers

However, the author doesn't go into much detail about most of these. In fact, it's almost as if he made a list of all the things he loved and wrote a book around it.  In some ways it feels a little like I was being manipulated, as if the author thought the mere mention of these subjects would stimulate some nostalgic pleasure center in my brain.

But, of course it did.

Daily Plaquenil Dosage - Crunching Numbers

NOTE: I've moved this blog post to my professional website, EyeDock. More importantly, I've written an update, Changes in Plaquenil Risk Calculations based on new recommendations based on a 2014 study. The calculator has also been updated to reflect these changes.

---

As all eye care providers should be aware, there have been some revisions in the recommendations of how we monitor our Plaquenil patients secondary to two articles1,2 published earlier this year.  

Whenever I want a quick review of Chloroquine and Hydroxychloroquine retinotoxicity I always head to Ron Melton and Randall Thomas's excellent website, Eyeupdate.  As expected, their site has been updated with the new recommendations and I won't bother duplicating their efforts here.

I will, however, plunge a little deeper into some of the dosaging concerns they brought up in their article.  After all, I never met an algebraic equation I didn't like, and I have a obsession for looking at data and trying to establish trends.  In addition, I thought I might be able to create a simple calculator to help my fellow eye care professionals determine their patient's retinotoxic risk as it relates to plaquenil dosage.  It's not that the calculations are particularly difficult, but, being Americans and using the moronic English system of measurements, it's requires a few more steps than a busy doctor should have to take. 

Going forward, I'm going to limit my discussion to hydroxychloroquine (Plaquenil) and women.  Chloroquine is less common, and men are less likely to be taking these medications.  Besides, the trends for men and Chloroquine will be very similar to what I discuss here.

Let's dive in.  Melton & Thomas describe the typical safe Plaquenil dosage this way:

The usual dosage is two 200mg tablets per day. This is generally a "safe" dose for patients having a lean body weight of at least 135 lbs. The generally-regarded safe dosage is 6.5mg/kg/qd, thus the 135 lb. safety threshold.

I think it's helpful to break down where this 135 lb threshold is coming from. First, let's convert the body weight to kilograms:

135 lbs x 0.45359237 Kg / lb = 61.2 kg

Then multiply the body weight and the "safe"daily dosage.

61.2 kg x 6.5mg/kg/qd = 398 mg/qd ---> ~two 200 mg tablets

To further visualize this, here it is in graphical form:

The blue line represents the generally-regarded safe daily dose per pt's weight (using the 6.5mg/kg/qd).  For individual's below 135 lbs the safe dosage falls below 400mg/qd, and, as a consequence, the typical dosage of two 200mg tablets carry a greater risk of retinotoxicity.

However, things get a little more complicated when we take into account obesity.  Again, from Melton & Thomas's Web site:

We stress here that HCQ is not absorbed into adipose tissue. This means that an obese person who weighs 160lbs may be considered within the "safe" zone, but their lean body weight may be only 120lbs, which poses an increased relative risk. We have communicated this critical issue of lean body weight to our referring rheumatologists and dermatologists, and have asked them to assess the lean body weight of each of their HCQ patients and make this information known to us, so that we can more accurately assess the risk for retinotoxicity . . .

I thought it would be helpful for my calculator to include the option of calculating lean body weight for overweight individuals.  The formulas for lean body weight are as follows:

Lean Body Weight (men) = (1.10 x Weight(kg)) - 128 ( Weight2/(100 x Height(m))2)

Lean Body Weight (women) = (1.07 x Weight(kg)) - 148 ( Weight2/(100 x Height(m))2)

I went ahead and built a calculator to 

1. take a patient's height and weight and calculate lean body weight
2. take the lean body weight and calculate the "safe" HCQ dosage

Unfortunately every set of numbers I plugged in indicated that my patient would be at a high risk dosage if they were taking 400 mg/day.  For example, let's take a 160 lb 5'4" woman:

160 lbs x 0.45359237 Kg / lb = 72.5 kg

5'4" = 64 inches  --> 64" * 0.0254 meters / inch = 1.625 meters

Lean body weight = (1.07 x 72.5kg) - 148 ( 72.5kg2/(100 x 1.625)2) = 48.2kg

48.2 kg x 6.5mg/kg/day = 313 mg / day

For this patient it appeared that the typical 400mg/day dosage would be too high.  I tried various other weights and heights and found that very few combinations resulted in a safe daily dosage above 400mg. Again, I decided to plot this:

After taking the patient's weight and using their height to calculate the lean mass, the "safe" daily dosages were calculated.  According to this graph, all but 6 foot tall 207lb women should be at risk for retinotoxic effects with 400mg daily dosages of plaquenil.  

As this is ridiculous, and has no relation to what we see in the real world, we've obviously made a bad assumption here.  If we define lean body weight as the body mass minus the body fat, well, we're always going to end up with a significantly smaller weight than a person's "real" weight, even for individuals at their ideal weight.  Multiplying this low number by our "safe" dosage will always result in a small recommended dosage, and obviously the 6.5mg/kg/qd wasn't meant for this.  Perhaps there is a conversion factor for safe dosage per kg of lean body weight, but, if there is, I haven't come across it.

This prompted me to re-read Revised Recommendations on Screening for Chloroquine and Hydroxychloroquine Retinopathy.2  This paper states:

Chloroquine and [Plaquenil] are not retained in fatty tissues, so patients who are obese could be seriously overdosed if medicated on the basis of weight alone.  Obese individuals should be dosed on the basis of height, which allows estimation of an asthenic or "ideal" body weight...

Aha - they're referencing ideal body weight, not lean body weight.  Ideal body weight is calculated with the following formulas:

Ideal Body Weight (men) = 50 + 2.3 ( Height(in) - 60 )

Ideal Body Weight (women) = 45.5 + 2.3 ( Height(in) - 60 )

As such, our 5'4" female would have an ideal body weight of:

5'4" = 64 inches

45.5 + 2.3 (64 - 60) = 54.7 kg

And a "safe" daily plaquenil dose of:

54.7 kg x 6.5mg/kg/qd = 356 mg / qd

So, for this patient it appeared that the typical 400mg/day dosage would still be too high.  Let's look at another graph:

Now we're comparing the patient's ideal weight, calculated from their height (note that this does not take into account the pt's true weight), and again calculating the "safe" daily dosage.  Interestingly, this indicates that a woman needs to be taller than 67 inches (5'7") before the 400mg/day dosage becomes a lower risk.  As the average woman is shorter than this height it would seemingly indicate that 400mg/day may be a high risk dose for the average woman.  However, this doesn't ring true as the typical dosage is 400mg/day and is considered to be appropriate for most individuals.

This made me curious about the disparity between dosing based on height (and calculating the ideal weight) or based solely on weight.  If every person is at their ideal weight it won't make any difference, but how about for overweight or obese individuals?  First, let's define "overweight" and "obese" - the Word Health Organization defines these terms based on the Body Mass Index (from Wikipedia):

BMI

Classification

< 18.5

underweight

18.5–24.9

normal weight

25.0–29.9

overweight

30.0–34.9

class I obesity

35.0–39.9

class II obesity

≥ 40.0

 class III obesity 

 ... where BMI = weight (kg) / height (m) 2

Let's take our 5'4" woman from above and see what she would weigh if she were "overweight" or "obese".  I'll arbitrarily choose the lower end of the range listed in the BMI table above (overweight BMI = 25).  So, we know that our 5'4" woman is 1.625 meters tall.  We'll plug that number into the above equation to see what weight would give us a BMI of 25:

25 = w / 1.6252

w = 25 * 1.6252 = 66.02 kg

Again, calculating "safe" daily plaquenil dosage for this weight:

66.02 kg * 6.5 mg/kg/qd = 429 mg

As expected, it indicates a higher "safe" dosage (429 vs 356mg) than the calculation using ideal body weight for this height. 

For the last time, let's graph this relationship to get a look at the big picture:

In this graph, the lowest line represents the ideal weight, which was calculated by the height.  The other two lines use the method above to calculate a weight that is equal to a BMI of 25 (overweight) or 30 (Obese I) for each height.  This effectively illustrates how assessing risk based on true weight alone would predict a much higher "safe" dosage than determining the acceptable dosage based on ideal weight (which, in turn, is based solely on the patient's height).

In the end what have we learned here?  Not very much, I'm afraid, at least not anything we didn't already know.  We obviously have to be aware of our patient's plaquenil dosage, especially in our overweight patients.  For overweight individuals it is helpful to use ideal body weight (based on height), rather than true weight, although it's not clear how overweight a person needs to before you switch formulas. Furthermore, ideal weight calculations seem to suggest a lower "safe" dosage than might be expected.  

Also, it's important to understand I've taken a couple general recommendations here and tried to extrapolate more information than I have a right to.   The Marmor2 article states:

A significant percentage of the reported HCQ toxic cases have been associated with daily doses > 6.5mg/kg . . .

which tells us 6.5 mg/kg is an appropriate guideline to follow but isn't necessarily a mathematical absolute, just like we know that not everyone who has an IOP above 24mmHg has glaucoma and that everyone below 18mmHg does not.

In addition, we cannot get too obsessed with daily dosages as other risk factors for retinotoxicity exist.  In fact, the cumulative dosage is thought to be a more significant risk factor, to the point where the incidence of HCQ retinotoxicity is less than 1% within the first 5 years of medication use.   Other risk factors include renal or liver disease, age, preexisting macular disease, and possibly genetic factors.

Despite the fact that 1) there may be some gray areas in Plaquenil dosage calculations and 2) daily dosage is not the only risk factor, it is a risk factor that the eye care specialist can assess.  Again, the calculations are not especially difficult, but I hope I can make them a little easier for the busy practitioner.

1 Michaelides, M., et al. "Retinal Toxicity Associated with Hydroxychloroquine and Chloroquine." Arch. Ophthalmic. Vol. 129, No. 1 January, 2011

2 Marmor, M. F., et al. "Revised Recommendations on Screening for Chloroquine and Hydroxychloroquine Retinopathy." Ophthalmology, Vol. 118, No. 2, February, 2011

A Second Display for $2.99

When my my workday is over and my kids go to bed I try to get a little exercise and then sit down at the kitchen table and start working on my laptop. Most of the time I'm either entering data or programming. Both of these tasks are cumbersome because I need to use a lot of different applications simultaneously. For example, while entering data I usually toggle between:

• A browser window open to www.eyedock.com
• A browser tab open to EyeDock's mySQL database admin page
• A browser tab or a PDF reader open to the data that I'm referring to 
• An FTP client
• An image editor
• A file manager

When programming I'm usually using:

• Whichever app I'm using to write my code (Flash, Xcode, or Textmate)
• An image editor 
• A browser to Google solutions to all the problems I run into
• The programming language documentation
• And, sometimes, another program to build my user interface (Interface builder)

I spend WAY too much time switching between tabs and windows. In addition, all the typing is starting to cause me some wrist and hand pain: I think the biggest problem is all the awkward motions on the laptop track pad trying to open, close, and shift windows, files, and palettes around.

DisplayPad iconI've started thinking I'd be much better off with a large second monitor. The problem is I don't really have an office area in my house. I do 95% of my work on the kitchen table, and about 5% in coffee shops. We're trying to buy a bigger house, so maybe I will have an office area at some point, but right now the priority is to give the kids their own rooms. And it's hard to justify buying a large second monitor when there's the uncertainty of how much we'll be spending on our mortgage in the near future.

I came across a very workable alternative after reading a blog post about an iPad app called DisplayPad. This app lets you use your iPad as an external monitor. All you have to do is download the app onto your iPad and put a free program on your Mac (downloaded from their website). In your Mac's System Preferences -> Displays you can tell the computer where the iPad is positioned in relation to you computer so you can easily move your cursor between them.System Preferences - > Displays

It may not seem like a 9.7" iPad screen would make that much difference, but it does. It actually increases my usable desktop area by about 50% (with a 15" MacBook Pro), which is significant.

It's not a perfect solution, though. The resolution isn't quite as sharp as my laptop monitor, and dragging windows or images around isn't very smooth. However, when I reserve my primary tasks for the main computer and delegate the iPad screen as a place to hold more static documents it works in a very satisfactory way.

I spent a big chunk of the day yesterday with my dual monitor system, and I found my workflow to be much smoother and efficient. Plus I only had to spend $2.99, and my second display is small enough that I could even take it with me to a coffee shop or on a trip if I needed to. It's a cheap solution . . . if you happen to have an iPad.

Dual displays in action