Fw: Glass-Based Radon Measurements

[Kai Kaletsch <info@eic.nu>]

Friends,



As a result of my request, a radsafer has sent me a copy of the Lagarde et

al. paper. I still seem to have a problem wrapping my head around this

case - control stuff. I would very much appreciate it if someone could take

the time to help me out (if you want to see the paper, you can get to a pdf

version from

http://www.nature.com/jea/journal/v12/n5/index.html ):



According to the first row of Table 4 of the Lagarde et al. paper there are

33 cases and 58 controls in the 50 to 80 Bq/m3 category ( i.e. 1.76 controls

per case) and 24 cases and 55 controls in the 80 to 140 Bq/m3 category (

i.e. 2.29 controls per case). Relatively speaking, there are 30 % more

healthy people (controls) in the higher dose category than in the lower dose

category. Yet, the authors calculate a higher relative risk for the 80 to

140 Bq/m3 category than in the 50 to 80 Bq/m3 category (1.42 vs. 1.28).



To me, that doesn't make any sense. If you asked me if I wanted to belong to

a group that is full of sick people or to a group that is full of healthy

people, I would answer: "I want to belong to the group that has lots of

healthy people, because that means my risk of becoming sick is lower."



Could someone please try to explain this in common sense terms?



Thanks in advance,

Kai



----- Original Message -----

From: "Kai Kaletsch" <info@eic.nu>

To: <epirad@mchsi.com>; <radsafe@list.vanderbilt.edu>

Sent: Saturday, August 31, 2002 9:42 AM

Subject: Re: Glass-Based Radon Measurements





> Bill and Friends,

>

> I'm still trying to wrap my head around what the actual observables are in

a

> case - control study. The way I understand it, case - control studies work

> like this:

>

> 1.    Find people with a disease (cases).

> 2.    Match people to the cases that are similar (controls).

> 3.    See if there is a difference in exposure to a suspected causal

agent.

> 4.    If there is a difference in exposure, you calculate an excess

relative

> risk (ERR) per unit exposure.

>

> The only observables are related to step 3, i.e. differences in exposure.

>

> My question is: Is there any way that we can infer the observables

> (differences in exposure) from the reported calculated results? (I don't

> have ready access to J Expo Anal Environ Epidemiol 2002 Sep;12(5):344-54)

>

> For example, could the statement:

>

> "...the excess relative risk (ERR) of 75% (-4% to 430%) per 100 Bq m(-3)

> obtained when using a continuous variable for surface-based average radon

> concentration estimates, were about twice the size of the corresponding

> relative risks obtained among these subjects when using air-based average

> radon concentration estimates."

>

> mean:

>

> "surface-based average radon concentration estimates showed only 1/2 as

much

> excess radon concentration of cases vs. controls than using air-based

> average radon concentration estimates."?

>

> In physics, the authors of studies usually put their raw data somewhere so

> that anyone can download it, re-analyze it, check for errors in arithmetic

> etc etc. Is this not usual in epidemiology? (Obviously one would have to

> hide the personal info of the participants, but that shouldn't be a

> problem.)

>

> To me, this would add more credibility to the studies. If the same people

> are designing the study and collecting the data and analyzing the data and

> writing conclusions about the data, would that not tend to amplify any

> biases these people might subconsciously have by a factor of 4?

>

> Best Regards,

> Kai

>

> ----- Original Message -----

> From: <epirad@mchsi.com>

> To: <radsafe@list.vanderbilt.edu>

> Sent: Thursday, August 29, 2002 3:14 PM

> Subject: Glass-Based Radon Measurements

>

>

> > Two recent papers that may be of interest.

> >

> > Regards, Bill Field

> > bill-field@uiowa.edu

> >

> > J Expo Anal Environ Epidemiol 2002 Sep;12(5):344-54

> >

> > Glass-based radon-exposure assessment and lung cancer

> > risk.

> >

> > Lagarde F, Falk R, Almren K, Nyberg F, Svensson H,

> > Pershagen G.

> >

> > Institute of Environmental Medicine, Karolinska

> > Institutet, Stockholm, Sweden.

> >

> > Lung cancer risk estimation in relation to residential

> > radon exposure remains uncertain, partly as a result of

> > imprecision in air-based retrospective radon-exposure

> > assessment in epidemiological studies. A recently

> > developed methodology provides estimates for past radon

> > concentrations and involves measurement of the surface

> > activity of a glass object that has been in a subject's

> > dwellings through the period for exposure assessment.

> > Such glass measurements were performed for 110 lung

> > cancer subjects, diagnosed 1985 to 1995, and for 231

> > control subjects, recruited in a case-control study of

> > residential radon and lung cancer among never-smokers in

> > Sweden. The relative risks (with 95% confidence

> > intervals) of lung cancer in relation to categories of

> > surface-based average domestic radon concentration

> > during three decades, delimited by cutpoints at 50, 80,

> > and 140 Bq m(-3), were 1.60 (0.8 to 3.4), 1.96 (0.9 to

> > 4.2), and 2.20 (0.9 to 5.6), respectively, with average

> > radon concentrations below 50 Bq m(-3) used as reference

> > category, and with adjustment for other risk factors.

> > These relative risks, and the excess relative risk (ERR)

> > of 75% (-4% to 430%) per 100 Bq m(-3) obtained when

> > using a continuous variable for surface-based average

> > radon concentration estimates, were about twice the size

> > of the corresponding relative risks obtained among these

> > subjects when using air-based average radon

> > concentration estimates. This suggests that surface-

> > based estimates may provide a more relevant exposure

> > proxy than air-based estimates for relating past radon

> > exposure to lung cancer risk.

> > --------------------------------------------

> >

> >

> > Health Phys 2002 Aug;83(2):261-71

> >

> >

> > 210Po implanted in glass surfaces by long term exposure

> > to indoor radon.

> >

> > Steck DJ, Alavanja MC, Field RW, Parkhurst MA, Bates DJ,

> > Mahaffey JA.

> >

> > Physics Department, St. John's University, Collegeville,

> > MN 56321, USA. dsteck@csbsju.edu

> >

> > Recent epidemiologic investigations of the relationship

> > between residential radon gas exposure and lung cancer

> > relied on contemporary radon gas measurements to

> > estimate past radon gas exposures. Significant

> > uncertainties in these exposure estimates can arise from

> > year-to-year variation of indoor radon concentrations

> > and subject mobility. Surface implanted 210Po has shown

> > potential for improving retrospective radon gas exposure

> > estimates. However, in previous studies, the ability of

> > implanted 210Po activity to reconstruct cumulative radon

> > gas exposure was not tested because glass was not

> > available from homes with known radon-gas concentration

> > histories. In this study, we tested the validity of the

> > retrospective radon gas reconstruction using implanted

> > 210Po surface activity by measuring glass surfaces from

> > homes whose annual-average radon gas concentrations had

> > been measured almost every year during two decades.

> > Regression analysis showed a higher correlation between

> > measured surface activity and cumulative radon gas

> > exposure in these homes (R2>0.8) than was observed in

> > homes where only contemporary radon gas measurements

> > were available. The regression slope (0.57 ky m(-1)) was

> > consistent with our earlier retrospective results.

> > Surface activity measurements were as reliable for

> > retrospective radon gas exposure reconstruction as

> > yearlong gas measurements. Both methods produced

> > estimates that were within 25% of the long-term average

> > radon gas concentrations in a home. Surface measurements

> > can be used for home screening tests because they can

> > provide rapid, reliable estimates of past radon gas

> > concentrations. Implanted 210Po measurements are also

> > useful in retrospective epidemiologic studies that

> > include participants who may have been exposed to highly

> > variable radon concentrations in previously occupied or

> > structurally modified homes.

> >

> >

> > ************************************************************************

> > You are currently subscribed to the Radsafe mailing list. To

unsubscribe,

> > send an e-mail to Majordomo@list.vanderbilt.edu  Put the text

"unsubscribe

> > radsafe" (no quote marks) in the body of the e-mail, with no subject

line.

> > You can view the Radsafe archives at http://www.vanderbilt.edu/radsafe/

> >

>



************************************************************************

You are currently subscribed to the Radsafe mailing list. To unsubscribe,

send an e-mail to Majordomo@list.vanderbilt.edu  Put the text "unsubscribe

radsafe" (no quote marks) in the body of the e-mail, with no subject line.

You can view the Radsafe archives at http://www.vanderbilt.edu/radsafe/