poorly controlled through randomization, has tremendous potential to influence occupational injury interventions and their wider generalization. Reduction of occupational injury intervention evaluation to one of statistical significance is irresponsible, and perhaps dangerous. “. . . it is . . . useful to take stock at regular intervals of who we are, where have we come from, and what has happened to our luggage.” 12 Hester J. Lipscomb, PhD John M. Dement, PhD, CIH Division of Occupational and Environmental Medicine Department of Community and Family Medicine Duke University Medical Center Durham, North Carolina E-mail: hester.lipscomb@duke.edu. References 1. Lehtola MM, van der Molen HF, Lappalainen J, et al. The effectiveness of interventions for preventing injuries in the construction industry, a system- atic review. Am J Prev Med 2008;35:77– 85. 2. Suruda A, Whitaker B, Bloswick D, Phillips P, Sesek R. Impact of the OSHA trench and excavation standard on fatal injury in the construction industry. J Occup Environ Med 2002;44:902–5. 3. Derr J, Forst L, Chen HY, Conroy L. Fatal falls in the U.S. construction industry, 1990 to 1999. J Occup Environ Med 2001;43:853– 60. 4. Lipscomb HJ, Li L, Dement JM. Work-related falls among carpenters in Washington State before and after the Vertical Fall Arrest Standard. Am J Indus Med 2003;44:157– 65. 5. Effective Practice and Organization of Care (EPOC). Criteria for systematic reviews. www.epoc.cochrane.org/Files/Website/Reviewer%20Resources/ inttime.pdf. 6. Kuhn L, Davidson LL, Durkin MS. Use of Poisson regression and time series analyses for detecting changes over time in rates of child injury following a prevention program. Am J Epid 1994;140:943–55. 7. Nelson N, Kaufman J, Kalat J, Silverstein B. Falls in construction: injury rates for OSHA-inspected employers before and after citation for violating the Washington State Fall Protection Standard. Am J Indus Med 1997;31:296 –302. 8. Zwerling C, Daltroy L, Fine LJ, Johnston JJ, Melius J, Silverstein B. Design and conduct of occupational injury intervention studies: a review of evaluation strategies. Amer J Indus Med 1997;32:164 –79. 9. Rychetnik L, Frommer M, Hawe P, Shiell A. Criteria for evaluating evidence on public health interventions. J Epidemiol Community Health 2002;56: 119 –27. 10. Vandenbroucke JP, von Elm E, Altman DG, et al. for the STROBE Initiative. Strengthening the reporting of observational studies in epidemiology (STROBE): explanation and elaboration. Epidemiology 2007;18:805–35. 11. Black N. Evidence based policy: proceed with care. Br Med J 2001;323: 275–9. 12. Pearce N. The rise and rise of corporate epidemiology and the narrowing of epidemiology’s vision. Int J Epid 2007;36:713–7. In Reply: We thank Drs. Lipscomb and Dement for their view on our Cochrane systematic review. We were glad to see that they were critical about the review, but also agreed with much of the content. The interpretation of interrupted time series is not straight- forward and can be easily subject to bias. In many studies, authors judge time trends purely based on eye-balling alone. Therefore we think that a statistical analysis in a standardized way will decrease the risk of bias. Another issue is when to expect the intervention to come into effect. To prevent data-driven analyses there should be a theoretical justification. In our approach, we assume sepa- rately an immediate effect and a long-term effect which has been used in many other reviews. 1 We did not have argu- ments to make a reasonable assumption about a lagging effect, so we analyzed all the studies in a standard way. We did our best in carefully combining the available studies, which is an important feature of Cochrane systematic reviews. This is why only the results of the three available time–series studies on regulation in the construction industry in the U.S. were combined. In our view, this was a considered choice that allows making inferences beyond the individual-study level. Even when we look at the data in context and do not condense it into a meta-analysis, we see three low-quality studies on legislation. Future interventions would certainly benefit from qualitative studies that can help in figuring out which barriers prevent the implementation of legislation. They can’t however replace the evidence of effectiveness, but they would be a welcome addition to the field of intervention effectiveness studies. The objective of our Cochrane systematic review was to gather evidence on the effectiveness of all available interventions in the construction industry, not just legislation. For interventions other than legislation, we did find evidence for effectiveness with the same methodology. However, this analysis was not chal- lenged at all by Lipscomb and Dement in their letter. Then again, even after lowering our inclusion criteria far below randomized studies and thorough searching, we found alto- gether only five studies that controlled for trends in time. We think that construction workers deserve better than that. We realize the difficulties in the evaluation of organiza- tional and regulatory interventions and can see that that may explain why so few effectiveness studies are available. We gathered also uncontrolled studies as part of our review, but we found it very difficult to draw conclusions from these as the authors of the studies themselves acknowledge. 2 We welcome and support all endeavors to do better here. We can only hope that Lipscomb and Dement agree that more and better studies are needed. We think that our review is an urgent call on all stakeholders to contribute more and better intervention-effectiveness studies, thus building up a valuable evidence base. We think that thorough implementa- tion of effective interventions is needed to save more lives in the construction industry. Marika M. Lehtola, MSc (Tech.) Finnish Institute of Occupational Health PO Box 93 (Neulaniementie 4), FI-70701 Kuopio, Finland E-mail: marika.lehtola@ttl.fi Henk F. Van der Molen, PhD Jorma Lappalainen, Lic.Sc. (Tech.) Peter L.T. Hoonakker, MSc Hongwei Hsiao, PhD Roger A. Haslam, PhD Andrew R. Hale, PhD Jos H. Verbeek, MD, PhD References 1. Ramsay CR, Matowe L, Grilli R, Grimshaw JM, Thomas RE. Interrupted time series design in health technology assessment: lessons from two systematic reviews of behavior change strategies. Int J Technol Assess Health Care 2003;19:613–23. 2. Darragh AR, Stallones L, Bigelow PL, Keefe TJ. Effectiveness of the HomeSafe pilot program in reducing injury rates among residential con- struction workers, 1994 –1998. Am J Ind Med 2004;45:210 –7. 378 American Journal of Preventive Medicine, Volume 36, Number 4 www.ajpm-online.net