Gas station without pumps

2019 January 20

Project Baseline claims (falsely) to prioritize sharing info with participants

Filed under: Uncategorized — gasstationwithoutpumps @ 11:56
Tags: , ,

In Blog – Project Baseline: A new era of health data: The vision and challenges of returning individual research results, the Project Baseline staff claim

At Project Baseline, we decided that we will prioritize the exploration of providing health information back to participants. We are well aware of the challenges linked to returning results but, in support of our philosophy that we want to take a patient-centric approach to clinical research, we are continuing to work through them.

Unfortunately, they are still not doing a very good job of sharing.

Some things they do well—for example, after the first visit they sent me the lab report on all the blood and urinalysis lab tests (which had no interesting results—the only anomalies were in the lipid test panel, which was flawed because it was not done fasting, and I already have lipid panels done annually).

Other things that they could provide trivially (like being able to download from the web portal the pulse-rate records that the watch collects) they do not provide at all.

Worst, though, is their practice of sharing only bad news.  For example, my December visit included a stress EKG and stress echocardiogram test, and some anomalies were found in the stress EKG.  Because stress EKG has a fairly high false positive rate, the usual followup is to do a stress echocardiogram, to see whether there is any cause for concern.  Their practice of sharing only bad news meant that they gave me the stress EKG results, but not the stress echocardiogram results (which they told me on the phone showed that the stress EKG is a false positive, but no written or e-mailed message said that).  I’m not even going to bother sharing the stress EKG results with my family physician, even though the letter they sent with the report says “We recommend that you contact your healthcare provider about these results”, because she will have no access to the stress echocardiogram that shows it to be a false positive and would be required to order a repeat of the test.

In a way, I’m glad that the stress EKG came up with the anomalies, because that meant that they let me have a copy of the 50-page EKG results, which they otherwise would not have.  I can show some of the EKG traces to my class, when they are making their own one-channel EKG amplifiers.

Also, the report confirmed my suspicion that we had not gotten anywhere near maximum effort on the stress test—my heart rate was still stepping up linearly with effort at the point the cardiologist ended the test, showing no signs of having plateaued. The cardiologist had convinced me to stop at only 6.8 METs (metabolic equivalents), while I routinely run an 8-minute mile, which is supposedly 11.8 METs [https://sites.google.com/site/compendiumofphysicalactivities/Activity-Categories/running].

If they do a stress echocardiogram at next year’s visit, I’m going ask that the cardiologist not stop me before 13 METs, 190bpm, or 220 mmHg systolic pressure, unless I first cry “hold, enough!”  I’d like to know what my max heart rate really is.  (If Project Baseline let me have access to my pulse-rate logs, I could see what my pulse rate gets up to on my daily bike commute or my jogging, which would at least give me a lower bound on my max heart rate.)

Update 2019 Jan 20:  I forgot to mention that some reports (it isn’t clear which) are available, according to an email message they sent:

Visit the Project Baseline mobile app to see your reports!

Please note that these are only available in the mobile app, not the web portal.

Unfortunately, because I don’t have a cellphone (much less a smartphone), I can’t see what is included in these reports.  I’m fairly certain that they don’t allow downloading the day’s or week’s record from the pulse monitor, though.

2018 December 16

Project Baseline

Filed under: Circuits course — gasstationwithoutpumps @ 10:58
Tags: , , ,

This month I joined a research project, Project Baseline, that is collecting detailed medical information on 10,000 subjects across the US.  I’ve not joined as a researcher, but as subject.

I wish I could tell you precisely what the goals of Project Baseline are, but none of their materials are very clear about that. The name suggests that they are trying to collect information about what results to expect from medical tests for asymptomatic patients—so many of the standard medical tests are only applied to people who are already ill that there is little idea what the results would be for a normal person.  Because the study runs for 4 years for each subject, they may also be looking for early-warning signals from tests—do any of the tests predict problems up to 4 years in advance?

The structure of the study is fairly straightforward—each subject gets extensive medical tests on an annual basis, wears a watch that monitors pulse and motion, and keeps a sleep monitor in their bed.  The selection of subjects is pretty broad: almost any US resident 18 or older who can get to one of the study sites is eligible.(You have not be allergic to nickel or metal jewelry, be able to speak and read English or Spanish, not work for Project Baseline, and be willing to share health information with them.)

I’ve been wanting to do a bunch of baseline fitness and health tests for a few years, but the retail cost of them has been prohibitive—with Project Baseline, not only were the tests free, but they pay me a tiny stipend for participating.  They also issued me a fancy watch for the pulse monitor and ECG and a sleep monitor to put under my mattress.

The watch, a Verily Life Sciences “Study Watch” has a nice E-ink display, but the firmware (BASELINE 000.048.009.BSL) is set up to make it less useful to me than my previous watch, a dime-store Casio model.  The study watch shows me only the date and time—and the time only to the nearest minute and only in 12-hour, not 24-hour format.  My dime-store watch showed seconds, could be set to 24-hour mode, and could be used as a stopwatch—none of which is possible with the study watch.

The wristwatch supposedly has an accelerometer for step counting—not a very useful feature for someone whose main aerobic exercise is bicycling. They should make an ankle bracelet for bicyclists—one that doesn’t look like the GPS ankle monitors worn by felons.  In any case, the step counter feature is not very useful, as the watch has no bluetooth synchronization, and only updates records when it is returned to its charging station.  That limitation makes it rather weird that the step count information is available only through a phone app, and not the Project Baseline web portal.  Because I don’t have a cellphone (and don’t see it as worthwhile to get one to view day-old step information), I have no access to this data record either.

The study watch is taking my pulse continually—it has a pair of green LEDs on the back of the watch that sample at 30 Hz (600µs on, 32.733 ms off).  It will also do a 1-minute ECG recording whenever I request one, which I’m supposed to do at least once a day.  Unfortunately, it never tells me my pulse nor is there anywhere I can look up either the pulse recordings or the ECG recordings, both of which I would find very interesting.  I’d really like to know what my pulse gets up to on my commute up the hill in the morning, and how long it takes to recover afterwards.  That information is being sent to Project Baseline, but they are not sharing it with me.

Similarly, the under-mattress sleep tracker is recording something about my sleep that I would be interested in knowing, but I have no access to that data either.

I had two 4–5-hour sessions of medical tests in Palo Alto as part of the study.  These seemed to focus mainly on cardiovascular tests, though there was a large panel of blood and urine tests also, and I had to ship them a bowl full of frozen poop by FedEx.

The first day included measuring my ankle-branchial index (the ratio of systolic blood pressure in the ankles to systolic blood pressure in the arms), to check for peripheral vascular disease.  One part of that is interesting, in that they listen to the blood flow with an ultrasonic Doppler-shift probe, which allows them to listen to a single artery at a time, presumably with more accuracy than just using a stethoscope.

They also took about 300 mL of my blood in many tiny tubes.  They ended up having to stick me three times: the first time the phlebotomist missed the vein, the second time they got only halfway through the tubes before the blood clotted enough to seal off the puncture, and the third time they managed to finish (though that puncture was the most painful of the lot).  I was not impressed with the competence of the phlebotomist—I’ve had less discomfort from the bigger needles of blood donations.

Other samples collected that day included urine, saliva (for oral microbiome I think, but possibly for genetic tests), tears, and swabs of ears, nose, inside cheeks.

They also took a 12-lead ECG, which I did not get to see.

They also spent a lot of time trying to get my medical history recorded.  I had brought detailed notes (which I left with them), but they had to enter it all into Google Forms, which seemed very poorly chosen for the task.

The second visit included some simple strength and balance tests, timed walking, vision screening, photos of cornea and retina, pulmonary function (including gas exchange), an echocardiogram, a stress EKG and echocardiogram, CT scan of my chest, and chest X-ray.  The timed walking consisted of some very short timed walks at low speed and high speed, then a 6-minute walk in the corridor at “normal speed”.  The person measuring that thought I might have set a course record in their hallway at 630m, but that is only 6.3 km/h or 3.9 mph, which not a particularly fast walking pace.  I suspect that many people find it difficult to walk at a normal pace when you have to reverse direction every 20m.

The pulmonary function and echocardiogram tests were the most interesting—it was particularly cool to see my heart beating and the false color showing the flow rate of the blood through the valves.

This is the MicroGard II mouthpiece that was used for the pulmonary tests.

The commercial mouthpiece used in the pulmonary function test might be useful for student lab kits next year, in place of the PVC elbow and barbed connector that the students currently use for their breath-pressure measurements.  The cost per student would be about the same, but I would have to make adapters as lab equipment—the circular tube fits tightly in a 1″ PVC slip connector, so it would not be difficult to make the adapters, but it would cost about $100 to make 25 adapters with barbed connectors and a 2mm vent hole.  The main advantages to using the commercial mouthpiece are that it has a filter to prevent cross-contamination, that it holds the mouth comfortably open, that it comes with a nose clip to prevent air leakage through the nose, and that it is a standard medical component, and so will seem more real to students.  The main disadvantage is that students would not get to practice using the drill press to drill the 2mm holes in the PVC elbows.  Of course, they would still get to drill holes in Lego bricks for the pulse-monitor lab.

The stress ECG and echocardiogram involved walking (and starting to jog) on a treadmill until the cardiologist thought I was working hard enough and had me immediately lie down to get an echocardiogram with my heart pounding.  I believe that the intent was to get me up to my maximum heart rate, but I don’t think we achieved that, as I was not feeling that I was going all out when the cardiologist started telling me to stop any time. I got my heart rate to 163 bpm (which is over the normal guideline of 220 − age in years), but I think I could have gone higher.  If they do the test again next year, I’ll have to tell the cardiologist not to hint at when to stop but to just keep increasing the treadmill speed until I call a halt.

They were supposed to administer a hearing test, but I don’t remember getting one.  It has been about a year since I last was tested by an audiologist, but I don’t think that there has been much change since then.  If they want the data, they can probably access my medical records.

Project Baseline has sent me a few results from the lab tests.  All the numerous blood and urinalysis standard lab tests came back in the normal range, except the cholesterol test—and I already knew about my cholesterol.  Besides which, the blood tests were not done fasting, so the cholesterol test was rather useless anyway—I did not bother adding it to my cholesterol records.

They also sent me some results from my CAT scan, which were again mainly normal: the biggest reports were that I had no spleen (which I knew already) and that I may have had a viral infection in my lung (I did have a cold that week).  The main thing they were looking for, calcification in the heart and arteries, came up with 0 calcium scores, though they also reported “trace focus of atherosclerotic calcification in the left anterior descending artery” and “calcifications of the aortic valve”.  Aortic valve calcification is a common condition in older adults, but “usually doesn’t cause significant heart problems, but requires regular checkups to make sure your condition isn’t worsening” according to the Mayo Clinic.  I suspect that Project Baseline will be doing those regular checkups for the next few years, but after the study ends I’ll have to make a point of having my valves checked every few years.

The chest X-rays did tell me something I didn’t know—I have “azygos fissure”, which occurs in 0.4–1.2% of the population (different sources give different frequencies, depending on how the estimate was made).  An azygos fissure is a result of a normal variant in embryonic development, in which the right lung grows around the right posterior cardinal vein, one of the precursors of the azygos vein, instead of the vein moving to the left as the lung grows. It generally is of no clinical significance, unless thoracic surgery is needed or it gets misinterpreted as something else when images are read.

%d bloggers like this: