Monday, May 5, 2008

Science: The Good, the Better, and the Junk

In the case of Conner Peterson's age, we see all three qualities of science: the good, the better, and the junk.

The Good

For decades, doctors relied on the first day of the last menstrual period to determine gestational age. The wheel was developed to predict the expected date of delivery (EDC) from this LMP. The expected delivery date was figured to be 40 weeks, and 2 weeks was tacked on at each end as a safety zone. The LMP for Laci determined that Conner's due date was February 10.

The Better

Then along came the ultrasound, also referred to as sonogram. In the first half of the pregnancy, the following are used:

7-9 weeks, crown-rump length provides maximum accuracy for gestational age
12-20 weeks, biparietal diameter provides maximum accuracy for gestational age

In the last half of the pregnancy, after 20 weeks, the 4 long bones -- humerus, tibia, ulna, and femur -- provide maximal accuracy. Jeanty, in his studies, measures all 4 bones and averages them to derive the gestational age.

Laci had 2 ultrasounds. The 1st one, performed on July 16, determined that Conner was 1 day younger than indicated by the LPM, which put his due date at February 11. Because it was so close to the LPM due date, the doctors did not note a corrected EDC.

The second one was performed on September 24. Since one of the purposes of performing a second ultrasound is to correct the EDC if necessary, Dr. Yip took 4 measurements:

Abdominal circumference (AC) 134 = 18w3d ± 14 days
Head circumference (HC) 163 = 19w0d ± 11 days
Biparietal diameter (BPD) 45 = 19w4d ± 10 days
Femur length (FL) 32 = 19w4d ± 6 days

He averaged them out and settled on 19w2d. This made Conner 6 days younger than the 1st ultrasound, and 7 days younger than the LMP. Dr. Yip noted February 16 as the corrected EDC on Laci's chart. In November and December, the notation for gestational age for each visit corresponds to corrected EDC.

Dr. Yip's conclusion is supported by fetal biometry studies. 10 studies produced 17 gestational ages for Conner, and only 1 of them confirmed the LPM and 1st ultrasound gestational ages. Only 1 out of 17. That one measurement was Jeanty's femur measurement. 3 other measurements by Jeanty produced gestational ages even younger than Yip's 19w2d, by as much as 8 days.

The LPM put Conner at 33w2d on December 24. The 1st ultrasound put him at 33w1d on December 24. The 2nd ultrasound put him at 32w2d on December 24. Jeanty's femur measurement (2nd ultrasound) put him at 33w1d.

Every eye witness that saw Conner when he was found described him as full term or 9-month, including Dr. Peterson who performed the autopsy. The one measurement that Peterson did aged Conner at 9 months.

Dr. Galloway, a forensic anthropologist, performed some measurements to determine Conner's age. Based on the three large bones (3 of the 4 that Jeanty recommends for the last half of the pregnancy), Conner's gestational age was 35-36 weeks.

35-36 weeks compared to these gestational ages on December 24:

LMP: 33w2d
1st ultrasound: 33w1d
Jeanty's femur (second ultrasound): 33w1d
2nd ultrasound: 32w2d

If you tack on the 2-week deviation, you come down to 33 weeks. But, don't you also have to tack the deviation on to the other measurements?

For example, the 1st ultrasound is considered accurate +/- 7 days. So, if you are going to use the bottom of the deviation to arrive at 33, don't you have to subtract 7 days from the 1st ultrasound age, making it 32w1d? Seems like that would be good science, if you use a deviation at one end of the spectrum, you have to use it at the other end also. Or you end up comparing apples to oranges.

But it is not reasonable to tack the standard deviation on, simply because every eye witness that saw Conner described him as a full-term or 9-month baby. If we were talking about the difference between 36 weeks and full term, the differences might not have been that noticeable, but a 33-week baby is noticeably different in appearance than a full-term baby. A 33-week baby, is at least 5 weeks premature, if you take 38-42 as full-term.

Even if you use Jeanty's gestational age for the femur on the 2nd ultrasound, you still have Conner too old to have died on December 24 -- 35.1 weeks from Sherwood compared to 33w1d -- that's a 2-week difference. And if you tack on the 2-week deviation to Sherwood, to derive 33, you also have to tack on the deviation to Jeanty.

The Junk

Enter Dr. Devore.

Devore ignored Galloway's measurements. She is a forensic anthropologist, trained in determining the ages of dead people, he is not.

Devore ignored Dr. Yip, who was Laci's doctor. Devore was not one of Laci's doctors.

Devore ignored the eye witnesses who saw Conner. He never saw Conner.

Devore ignored Jeanty, who says that the biparietal measurement is the most accurate from 12-20 weeks.

Devore ignored 16 fetal measurements from 10 fetal biometry studies that put Conner at a younger age than both the LMP and the 1st ultrasound.

Devore ignored 3 out of the 4 measurements per Jeanty's studies -- all 3 which produced younger ages for Conner.

Devore used an ultrasound method never before used to determine fetal age -- he took a bone that had been removed from a dead fetus and compared it to the measurements of the fetus when it was alive and still in the mother's womb. His method has never been subjected to peer review.

The femur bone he used had been out of the body from April 14-February 2, 9 1/2 months. We don't know under what conditions it was stored. We don't know how much it dried. Have you ever heard the expression, dry as a bone, or bone-dry? There's a reason for that expression. We would know if it shrunk any if Devore had bothered to measure it the same way Galloway did. Then we would know with certainty if it was still the same length as when it was removed from Conner's body. But, he didn't do any of the quality control checks that good science does.

Devore ignored the other 2 large bones, the humerus and the tibia, that were available to him, which Galloway had to do her measurements. Jeanty says 4 large bones must be used and averaged, that only one bone is bound to produce error. If he would have used 3 of the 4, then averaged them, that would have been much closer to what Jeanty did in his studies. But he didn't do that, or at least he didn't admit he did.

Devore's first calculation was December 25 -- which he said was wrong because he "counted on his fingers." Good scientists don't count on their fingers when doing "expert opinion" reports for a court of law when a man's life is on the line.

Devore may be good at his profession, but he used junk science to help convict Scott Peterson. If Devore can't tell the difference between good science and junk science, then he shouldn't be practicing medicine. If he can tell the difference, and did it anyway, shouldn't he receive the same punishment he helped to impose on Scott? At the very least, he should serve the same amount of prison time, in the same prison and under the same conditions, as Scott ends up serving before he is exonerated. Scott has already been in jail/prison for 5 years.

And what about the added pain he is going to cause the family when they are forced to admit that Scott was not responsible for Laci's abduction and he did not murder Laci and Conner. Whatever hell Sharon Rocha and others went through the first time around, it is going to be even worse the second time.


We are no longer accepting comments on the Blog. To respond to this article, please use our SII Yahoo! Group,