This article has Open Peer Review reports available.
Assessing regression to the mean effects in health care initiatives
© Linden; licensee BioMed Central Ltd. 2013
Received: 2 July 2013
Accepted: 20 September 2013
Published: 28 September 2013
Interventions targeting individuals classified as “high-risk” have become common-place in health care. High-risk may represent outlier values on utilization, cost, or clinical measures. Typically, such individuals are invited to participate in an intervention intended to reduce their level of risk, and after a period of time, a follow-up measurement is taken. However, individuals initially identified by their outlier values will likely have lower values on re-measurement in the absence of an intervention. This statistical phenomenon is known as “regression to the mean” (RTM) and often leads to an inaccurate conclusion that the intervention caused the effect. Concerns about RTM are rarely raised in connection with most health care interventions, and it is uncommon to find evaluators who estimate its effect. This may be due to lack of awareness, cognitive biases that may cause people to systematically misinterpret RTM effects by creating (erroneous) explanations to account for it, or by design.
In this paper, the author fully describes the RTM phenomenon, and tests the accuracy of the traditional approach in calculating RTM assuming normality, using normally distributed data from a Monte Carlo simulation and skewed data from a control group in a pre-post evaluation of a health intervention. Confidence intervals are generated around the traditional RTM calculation to provide more insight into the potential magnitude of the bias introduced by RTM. Finally, suggestions are offered for designing interventions and evaluations to mitigate the effects of RTM.
On multivariate normal data, the calculated RTM estimates are identical to true estimates. As expected, when using skewed data the calculated method underestimated the true RTM effect. Confidence intervals provide helpful guidance on the magnitude of the RTM effect.
Decision-makers should always consider RTM to be a viable explanation of the observed change in an outcome in a pre-post study, and evaluators of health care initiatives should always take the appropriate steps to estimate the magnitude of the effect and control for it when possible. Regardless of the cause, failure to address RTM may result in wasteful pursuit of ineffective interventions, both at the organizational level and at the policy level.
Interventions targeting individuals classified as “high-risk” have become common-place in the health care industry. High-risk may capture anything from high utilization or cost of health services, to outlier values on clinical measures (e.g., blood glucose, blood pressure, cholesterol). Typically, such individuals are invited to participate in an intervention intended to reduce their level of risk, and after a period of time, a follow-up measurement is taken. The pre-test to post-test change in the outcome is then generally presented as the impact of the intervention. This evaluation approach is problematic from a statistical standpoint because individuals initially identified by their high values will likely have lower values on re-measurement in the absence of an intervention. This statistical phenomenon is known as “regression to the mean” (RTM) and often leads to an inaccurate conclusion that the intervention resulted in a treatment effect .
The implications of RTM in evaluating medical interventions have been examined extensively in the literature, suggesting that RTM is a common problem [2–21]. However, RTM is rarely addressed when evaluating health care delivery interventions or in the more general decision making processes in health care . This is despite its increasing relevance given the intensified focus on high-cost, and high-need groups, and efforts to design programs specifically targeting them. There are at least three possible explanations for why this may be. First, evaluations of delivery-side interventions are traditionally not subject to the same rigor as medical interventions (i.e., RCTs), and with this may come a lack of awareness of the need to address RTM. Second, it has been shown that cognitive biases may cause people to subconsciously systematically misinterpret RTM effects as intervention effects by creating (erroneous) explanations to account for it . Third, there are more blatant examples in which organizations have a stake in the outcome of the intervention and capitalize on the RTM effect as a business strategy. For example, commercial disease management organizations have long advocated that their programs be evaluated without a control group despite the recognition that the intervention group will demonstrate better outcomes due to regression to the mean . Regardless of the cause, failure to address RTM may result in wasteful pursuit of ineffective interventions, both at the organizational level and at the policy level.
In this paper, we seek to provide researchers, organizational decision-makers, and policy-makers, with a broa der set of tools to understand and assess RTM effects. First, real examples of RTM in health care are presented to illustrate the phenomenon. Next, the traditional method for calculating the RTM effect in normally distributed data is described, and these RTM effect estimates are compared with RTM effects generated from Monte Carlo simulation of normally distributed data. Next, these comparisons are repeated using skewed data from a control group in a health coaching study to illustrate the shortcomings of the traditional approach to accurately estimate the RTM effect, in the common scenario of non-normal data. We use this to motivate the primary contribution of the paper, the estimation of standard errors and confidence intervals around the RTM effect. While largely absent from existing RTM literature, including confidence intervals, a measure of the precision of single-value RTM estimates, is valuable because it provides a range of values that are considered to be plausible for the population. Finally, the advantage of calculating confidence intervals around RTM estimates is discussed in detail, and approaches for designing health care interventions to mitigate, or at least account for, the effects of RTM are provided.
The regression to the mean concept
Regression to the mean was first described over a century ago by Francis Galton (later Sir Francis) upon discovering that, on average, tall parents have children shorter than themselves and short parents have taller children than themselves . RTM is the result of both random measurement error and extremity of scores from the mean . A simple example of this occurs in measuring blood pressure or heart rate. Rarely are any two observations identical, even if taken minutes apart, due to natural biologic variability or measurement error. At the individual level this is called within-subject variability. Additionally, the more extreme the initial value, the greater the expected change will be in the follow-up score. However, over the course of many repeated observations, this variability narrows around the true mean [27, 28]. Similar to individual level measures, groups with high (or low) initial mean values will also tend to regress to the mean of the overall sample.
In the context of an intervention, RTM can easily be mistaken for a program effect in the absence of an equivalent comparison group. The best approaches to illustrate the RTM phenomenon are either by using observations taken from time periods in which no interventions were implemented, or by using control group data derived from a research study.
The examples presented in this section use measures that commonly serve as outcomes in health care interventions, and both cases clearly illustrate RTM. This suggests that there are likely many contexts in which RTM, and not a program effect, explains an observed change from initial outlier status to follow-up values closer to the overall mean.
Classic formulae for estimating the magnitude of RTM
where Φ(z) is the probability density function and Φ(z) is the cumulative distribution function for z in a standard normal distribution.
where values are added whenever high-risk is indicated by values above κ, and subtracted whenever high-risk is indicated by values below κ. Subtracting the expected pre-test mean (Equation 5) from the expected post-test mean (Equation 6) should elicit the same expected RTM effect as that derived in Equation 1 (as will Cσ(1-ρ)).
Testing the performance of the RTM formulae
We examine the performance of the RTM equation (Equation 1) in estimating the RTM effect using two approaches. First a Monte Carlo simulation study is conducted assuming medical cost as the outcome, as it is often a primary focus of health services research. Following the simulation, the performance of the RTM formulae is demonstrated using actual data (the PCS data described in the previous section).
Design of the Monte Carlo simulation
An “actual” RTM effect is generated by drawing two variables from a multivariate normal distribution to represent the pre-test and post-test costs for a pseudo-population of 10,000 observations, with means of $5,000, standard deviations of $1,350, and three pretest-posttest correlations: 0.25, 0.50 and 0.75. The minimum value of the highest pre-test quintile of cost is set as the cutoff (≈ $6,136), with values above and below that level categorized as “high-risk” and “low-risk”, respectively. The mean difference in pretest-posttest costs for the two risk tiers represents the “actual” RTM effects. We compare this with the “calculated” RTM effect for the high- and low-risk groups using Equation 1 with the same cutoff value (≈ $6,136). This process is repeated 10,000 times for each of the three correlation levels and the actual versus calculated RTM effects are reported for the low and high-risk groups. The simulation was conducted in Stata 12.1 (StataCorp, College Station, TX), using the built in simulate command, and rtmci, a command written by the author (available upon request).
Design of the empirical example
Here, the PCS data for the 118 controls  described earlier in the current paper and illustrated in Figure 1 are revisited, in order to demonstrate the performance of the RTM formulae when data are skewed (p<0.00001 for the Shapiro-Wilk W test). The pre-test mean, post-test mean, and mean difference in pretest-posttest PCS scores are used to generate the “actual” RTM effects and the “calculated” comparisons are again computed using Equations 1, 5 and 6 and rtmci in Stata. The differences between the actual and calculated values are then compared, and for all estimates, 95% confidence intervals are computed via bootstrap simulation, e.g., by resampling 1000 observations (with replacement) from the actual data.
Monte Carlo simulation
Results of the Monte Carlo simulation ( N = 10,000)
[95% Confidence Interval]
ρ = 0.25
RTM (H) actual
RTM (H) calculated
RTM (L) actual
RTM (L) calculated
ρ = 0.50
RTM (H) actual
RTM (H) calculated
RTM (L) actual
RTM (L) calculated
ρ = 0.75
RTM (H) actual
RTM (H) calculated
RTM (L) actual
RTM (L) calculated
Regression to the mean effects for Physical Component Summary (PCS) scores on the Short Form-12 (SF-12v2), from the high-risk (PCS values ≤ 44.25) subgroup of controls ( n = 34) participating in a health coaching study (Butterworth et al. 2006)
[95% Confidence Interval]
Difference (Actual – Calculated)
The results of the simulation study demonstrate that the formulae for estimating RTM effects [27, 33] accurately calculate RTM when the data are normally distributed. By extension, these results support the use of RTM analysis in pre-post observational studies as a means of estimating the RTM effect. However the results using these skewed data, suggests the RTM calculation significantly under-estimated the true RTM effect by between 1.12 and 8.68 points. Generally, when researchers seek to calculate the RTM effect using skewed data, transforming the data to make them normally distributed before using the traditional formulae may suffice. However, if transforming data to another scale may lead to a loss of interpretability (as would be the case with SF-12 data), performing the calculations on the original scale and calculating confidence intervals that reveal the magnitude of the error offers an alternative approach that may be more useful. In our example, the confidence interval for the calculated RTM effect was 1.54 to 5.22, which overlaps with the actual RTM confidence interval of 4.35 to 12.21. Thus, the confidence interval for the calculated RTM effect provides a range of values more closely aligned with the true effect than the point estimate alone. A third option is to consider models devised to estimate regression to the mean effects in non-normally distributed data [35, 36]. However, some of these approaches rely on non-parametric modeling approaches, such as kernel density estimators , and are sensitive to the choice of bandwidth. Thus, the various approaches to estimating RTM may likely elicit different estimates depending on which methods are employed, even within the same data-set. Here again, the addition of confidence intervals can provide assistance to the evaluator in determining the overlap in estimates derived among the various methods.
Designing interventions to mitigate the RTM effect
While earlier sections focused on illustrating RTM and offering suggestions for how to estimate the magnitude of the RTM effect, ideally studies are designed to mitigate the effect of RTM. The randomized-controlled trial (RCT) is the obvious study design to control for RTM because randomly assigned groups should be equally affected (i.e., the treatment effect is the net effect after eliminating any RTM). The regression-discontinuity (RD) design should be considered as a viable alternative when randomization is not possible [37, 38]. The RD design relies on a cut-off point on a continuous pre-intervention variable to assign individuals to treatment. The individuals just to the right and left of the cutoff are assumed to be exchangeable - as in a randomized trial. Because individuals do not have precise control over their assignment score (nor would they know where the cutoff lies), they cannot self-select into treatment. Thus, we would expect a similar RTM effect for both groups in the neighborhood of the cutoff.
A third approach to mitigating the RTM effect in the design stage of an intervention is to base the treatment assignment on the cut-off, conditioned on the mean of multiple pre-tests rather than just a single pre-test [27, 33, 39, 40]. This has the effect of stabilizing the mean and reducing within-subject variability. When multiple pre-test measurements are used, the previously described equations require minor modification [27, 28, 33]. In Equation 1, the within-subject variability y 2 (in both the numerator and denominator) is now divided by the number of pre-tests n from which the mean is derived, becoming y 2 /n. In all other equations, σ is now replaced with the pooled standard deviation adjusted for multiple pre-test periods .
Controlling for RTM through data analysis
When only retrospective observational data are available, several approaches may be considered to control for RTM. Matching techniques  allow the investigator to try to replicate the randomization process by creating a control group that is essentially equivalent to the treatment group on observed pre-intervention characteristics – especially on the pre-test variable that we are most concerned leads to RTM. One particular advantage of matching techniques over other covariate adjustment strategies (e.g., multiple regression models), is that the investigator can directly assess how well the pre-test variable overlaps in its distribution between groups using graphical or numerical diagnostics . A high degree of overlap in the distribution increases our confidence that the RTM is effectively controlled for, as we would expect in an RCT.
where is the treatment group mean, ρ is the pre-post correlation for that treatment group, and x is the individual’s pre-test value. It is important to keep in mind, however, that when using ANCOVA (with or without the corrected pre-test), model assumptions, such as linearity between outcome and covariates, must be tested. Moreover, contrary to matching strategies where covariate balance can be directly assessed, in ANCOVA models, there is no assurance that the treatment groups are comparable on all baseline covariates. In fact, it is imperative that decision-makers consider other potential sources of bias (in addition to RTM) that may masquerade as a treatment effect. This is particularly true when using observational data, since it is likely that participants and non-participants will differ on several characteristics (e.g., health behaviors) not often available in claims analysis .
Finally, perhaps the easiest approach for adjusting outcomes to control for RTM effects is simply to subtract the calculated RTM effect derived from Equation 5 from the overall treatment effect estimate . Moreover, with the additional availability of confidence intervals, the investigator can provide a range of “net” treatment effect estimates when data are skewed.
In this paper we have illustrated that health care interventions are susceptible to the effects of RTM when individuals are chosen to participate in the intervention based on their outlier baseline “risk” score, and there is large within-subject variability or measurement error. When estimating the RTM effect on normally distributed data the calculated estimates produce identical results to those of simulated data. However, the equations underestimated the RTM effect in right-skewed data. We described several approaches for investigators to consider as methods to adjust for RTM, depending on the degree of control they have over the intervention and evaluation designs. However, designing interventions to mitigate the effects of RTM is a preferred strategy to retrospectively estimating the extent to which RTM may explain any observed treatment effect. Most importantly, both evaluators and stakeholders should be aware of RTM as a major source of bias in intervention studies, and take the appropriate steps to estimating its effect and controlling for it whenever possible to ensure valid conclusions about program effectiveness.
All of the proposed methods in this paper can be implemented via companion software, for Stata, which was written by the author and is available upon request.
I wish to thank Nicholas J. Cox for his extremely helpful comments while developing the companion software package and for reviewing the manuscript, and to Julia Adler-Milstein for her excellent review and comments on the manuscript.
- Bland JM, Altman DG: Regression towards the mean. BMJ. 1994, 308: 1499-10.1136/bmj.308.6942.1499.View ArticlePubMedPubMed CentralGoogle Scholar
- Boissel JP, Duperat B, Leizorowicz A: The phenomenon of regression to the mean and clinical investigation of blood cholesterol lowering drugs. Eur J Clin Pharmacol. 1980, 17: 227-230. 10.1007/BF00561905.View ArticlePubMedGoogle Scholar
- Andrews G, Harvey R: Regression to the mean in pretreatment measures of stuttering. J Speech Hear Disord. 1981, 46: 204-207.View ArticlePubMedGoogle Scholar
- Shepard DS, Finison LJ: Blood pressure reductions: correcting for regression to the mean. Prev Med. 1983, 12: 304-317. 10.1016/0091-7435(83)90239-6.View ArticlePubMedGoogle Scholar
- Whitney CW, Von Korff M: Regression to the mean in treated versus untreated chronic pain. Pain. 1992, 5: 281-285.View ArticleGoogle Scholar
- Denke MA, Frantz ID: Responses to a cholesterol-lowering diet: efficacy is greater in hypercholesterolemic subjects even after adjustment for regression to the mean. Am J Med. 1993, 94: 626-631. 10.1016/0002-9343(93)90215-B.View ArticlePubMedGoogle Scholar
- Herpin D, Demange J: Effect of regression to the mean in serial echocardiographic measurements of left ventricular mass: quantification and clinical implications. Am J Hypertens. 1994, 7: 824-828.PubMedGoogle Scholar
- Forrow L, Calkins DR, Allshouse K, Horowitz G, Delbanco TL: Evaluating cholesterol screening: the importance of controlling for regression to the mean. Arch Intern Med. 1995, 155: 2177-84. 10.1001/archinte.1995.00430200056009.View ArticlePubMedGoogle Scholar
- Prescott RJ, Garraway WM: Regression to the mean occurs in measuring peak urinary flow. Br J Urol. 1995, 76: 611-613. 10.1111/j.1464-410X.1995.tb07787.x.View ArticlePubMedGoogle Scholar
- Pitts SR, Adams RP: Emergency department hypertension and regression to the mean. Ann Emerg Med. 1998, 31: 214-8. 10.1016/S0196-0644(98)70309-9.View ArticlePubMedGoogle Scholar
- Murakawa Y, Yamashita T, Ajiki K, Sezaki K, Omata M: Ostensible day-night difference of QT prolongation during long-term treatment with antiarrhythmic drugs: reappraisal of the law of ‘regression to the mean’. J Cardiovasc Pharmacol. 1998, 32: 62-65. 10.1097/00005344-199807000-00010.View ArticlePubMedGoogle Scholar
- Cummings SR, Palermo L, Browner W, Marcus R, Wallace R, Pearson J, Blackwell T, Eckert S, Black D: Monitoring osteoporosis therapy with bone densitometry: misleading changes and regression to the mean. JAMA. 2000, 283: 1318-1321. 10.1001/jama.283.10.1318.View ArticlePubMedGoogle Scholar
- Asmar R, Safar M, Oueneau P: Evaluation of the placebo effect and reproducibility of blood pressure measurement in hypertension. Am J Hypertens. 2001, 14 (6 Pt 1): 546-552.View ArticlePubMedGoogle Scholar
- Chapurlat RD, Blackwell T, Bauer DC, Cummings SR: Changes in biochemical markers of bone turnover in women treated with Raloxifene: influence of regression to the mean. Osteoporosis Int. 2001, 12: 1006-1014. 10.1007/s001980170009.View ArticleGoogle Scholar
- Takashima Y, Sumiya Y, Kokaze A, Yoshida M, Ishikawa M, Sekine Y, Akamatsu S: Magnitude of the regression to the mean within one-year intra-individual changes in serum lipid levels among Japanese male workers. J Epidemiol. 2001, 11: 61-69. 10.2188/jea.11.61.View ArticlePubMedGoogle Scholar
- Krum H, Tonkin A: Why do phase III trials of promising heart failure drugs often fail? The contribution of ‘regression to the truth’. J Card Fail. 2003, 9: 347-349. 10.1054/j.cardfail.2003.08.006.View ArticleGoogle Scholar
- Allison DB, Loebel AD, Lombardo I, Romano SJ, Siu CO: Understanding the relationship between baseline BMI and subsequent weight change in antipsychotic trials: Effect modification or regression to the mean?. Psychiatry Res. 2009, 170: 172-176. 10.1016/j.psychres.2008.10.007.View ArticlePubMedGoogle Scholar
- Aaron MT, Applegate RA, Porter J, Thibos LN, Schallhorn SC, Brunstetter TJ, Tanzer DJ: Why Preoperative Acuity Predicts Postoperative Acuity in Wavefront-Guided LASIK. Optom Vis Sci. 2010, 87: 861-866. 10.1097/OPX.0b013e3181f6fb49.View ArticlePubMedPubMed CentralGoogle Scholar
- Bernstein JA, Bernstein E, Heeren TC: Mechanisms of change in control group drinking in clinical trials of brief alcohol intervention: Implications for bias toward the null. Drug Alcohol Rev. 2010, 29: 498-507. 10.1111/j.1465-3362.2010.00174.x.View ArticlePubMedGoogle Scholar
- Taylor CE, Jones H, Zaregarizi M, Cable NT, George KP, Atkinson G: Blood pressure status and post-exercise hypotension: an example of spurious correlation in hypertension research?. J Hum Hypertens. 2010, 24: 585-592. 10.1038/jhh.2009.112.View ArticlePubMedGoogle Scholar
- McCall WV, D’Agostino R, Rosenquist PB, Kimball J, Boggs N, Lasater B, Blocker J: Dissection of the factors driving the placebo effect in hypnotic treatment of depressed insomniacs. Sleep Med. 2011, 12: 557-564. 10.1016/j.sleep.2011.03.008.View ArticlePubMedPubMed CentralGoogle Scholar
- Morton V, Torgerson DJ: Effect of regression of the mean in decision making in health care. BMJ. 2003, 326: 1083-1084. 10.1136/bmj.326.7398.1083.View ArticlePubMedPubMed CentralGoogle Scholar
- Tversky A, Kahneman D: Judgment under uncertainty: Heuristics and biases. Science. 1974, 185: 1124-1131. 10.1126/science.185.4157.1124.View ArticlePubMedGoogle Scholar
- Linden A: Use of the total population approach to measure U.S. disease management industry's cost savings: issues and implications. Dis Manage Health Outc. 2007, 15: 13-18. 10.2165/00115677-200715010-00003.View ArticleGoogle Scholar
- Galton F: Regression towards mediocrity in hereditary stature. J Anthropol Inst. 1886, 15: 246-263.Google Scholar
- Campbell DT, Kenny DA: A Primer on Regression Artifacts. 1999, New York: Guilford PressGoogle Scholar
- Gardner MJ, Heady JA: Some effects of within person variability in epidemiological studies. J Chronic Dis. 1973, 26: 781-795. 10.1016/0021-9681(73)90013-1.View ArticleGoogle Scholar
- Johnson WD, George VT: Effect of regression to the mean in the presence of within-subject variability. Stat Med. 1991, 10: 1295-1302. 10.1002/sim.4780100812.View ArticlePubMedGoogle Scholar
- Linden A, Adams J, Roberts N: An assessment of the total population approach for evaluating disease management program effectiveness. Dis Manage. 2003, 6: 93-102. 10.1089/109350703321908478.View ArticleGoogle Scholar
- Ware JE, Kosinski M, Keller SD: A 12-item short-form health survey: Construction of scales and preliminary tests of reliability and validity. Med Care. 1996, 34: 220-233. 10.1097/00005650-199603000-00003.View ArticlePubMedGoogle Scholar
- Butterworth S, Linden A, McClay M, Leo MC: The effect of motivational interviewing-based health coaching on employees' physical and mental health status. J Occup Healt Psych. 2006, 11: 358-365.View ArticleGoogle Scholar
- Ware JE, Losinski M, Turner-Bowker DM, Gandek B: How to Score Version 2 of the SF-12® Health Survey (With a Suppl. Documenting Version 1). 2002, Lincoln, RI: QualityMetric IncorporatedGoogle Scholar
- Davis CE: The effect of regression to the mean in epidemiologic and clinical studies. Am J Epidemiol. 1976, 104: 493-498.PubMedGoogle Scholar
- Barnett AG, van der Pols JC, Dobson AJ: Regression to the mean: what it is and how to deal with it. Int J Epidemiol. 2005, 34: 215-220.View ArticlePubMedGoogle Scholar
- Chesher A: Non-normal variation and regression to the mean. Stat Methods Med Res. 1997, 6: 147-166. 10.1191/096228097672663908.View ArticlePubMedGoogle Scholar
- Majnu J, Abbas FJ: Assessing the regression to the mean for non-normal populations via kernel estimators. N Am J Med Sci. 2010, 2: 288-292.Google Scholar
- Lee DS, Lemieux T: Regression discontinuity designs in econometrics. J Econ Lit. 2010, 48: 281-355. 10.1257/jel.48.2.281.View ArticleGoogle Scholar
- Linden A, Adams JL: Combining the regression-discontinuity design and propensity-score based weighting to improve causal inference in program evaluation. J Eval Clin Pract. 2012, 18: 317-325. 10.1111/j.1365-2753.2011.01768.x.View ArticlePubMedGoogle Scholar
- Yudkin PL, Stratton IM: How to deal with regression to the mean in intervention studies. Lancet. 1996, 347: 241-243. 10.1016/S0140-6736(96)90410-9.View ArticlePubMedGoogle Scholar
- Linden A: Estimating the effect of regression to the mean in health management programs. Dis Manage and Healt Outc. 2007, 15: 7-12. 10.2165/00115677-200715010-00002.View ArticleGoogle Scholar
- Stuart EA: Matching methods for causal inference: a review and a look forward. Stat Sci. 2010, 25: 1-21. 10.1214/09-STS313.View ArticlePubMedPubMed CentralGoogle Scholar
- Roberts AOH: Regression toward the mean and the regression-effect bias. New Directions for Testing and Measurement. Edited by: Echternacht G. 1980, San Francisco: Jossey-Bass, 59-82.Google Scholar
- Irwig L, Glasziou P, Wilson A, Macaskill P: Estimating an individual’s true cholesterol level and response to intervention. JAMA. 1991, 266: 1678-85. 10.1001/jama.1991.03470120080037.View ArticlePubMedGoogle Scholar
- Trochim WMK: The Research Methods Knowledge Base. 2001, Cincinnati: Ohio, Atomic Dog Publishing, 2Google Scholar
- Shadish SR, Cook TD, Campbell DT: Experimental and quasi-experimental designs for generalized causal inference. 2002, Boston: Houghton MifflinGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2288/13/119/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.