This paper describes schools' capacity to implement non-pharmaceutical interventions during the pandemic influenza (specifically pH1N1) among school-aged children by a large public school system in the United States. The NYCDOE is the largest system of public schools in the United States, serving about 1.1 million students in nearly 1,700 schools 
. Our survey findings suggest that many public schools implemented many of the recommended NPIs by the NYC health and school officials. During the 2009 fall term, nearly all respondents reported teaching curriculum on proper hand hygiene and respiratory etiquette. Implementation of other NPIs was variable.
Planning for an influenza outbreak in public schools was one of the hallmarks of DOHMH mitigation efforts against pH1N1. The capacity of OSH in getting schools to implement the guidelines cannot be fully assessed based on this evaluation, but evidently a high percentage of the survey participants implemented recommendations for planning for an influenza outbreak. Two key recommendations in the DOHMH planning guidelines are formation of a Flu Response Team as a part of school emergency preparedness plan, composed of school administrators, health officials, and parents, and designation of a holding room within a school to be used exclusively for separating persons with ILI symptoms. The majority of respondents, but not all, had a Flu Response Team and a holding room. The barriers to adoption of these key recommendations are unclear, but schools without an FRT were more likely to serve older students and had a smaller percentage of white students than those with an FRT. Moreover, schools with an FRT were more likely to implement more aspects of the mitigation guidelines, including isolating students with ILI symptoms, than were schools without an FRT, highlighting the importance of planning as a significant step in implementation of the mitigation guidelines. It is possible that many schools that did not respond to the survey did not have an FRT as part of their required emergency preparedness plans. It is also possible that these schools were less likely to implement the mitigation guidelines. However, other studies that looked at the use of NPIs to limit the spread of pH1N1 in schools revealed that many schools and universities in other parts of the United States adopted most CDC-recommended NPIs but compliance with certain NPIs, especially isolating students with ILI symptoms, was low 
. These findings underscore the need to provide feasible recommendations that incorporate individual school needs and to allocate resources to address barriers to planning for influenza and other respiratory disease outbreaks and adoption of mitigation guidelines. Barriers to planning for pH1N1 outbreak by individual schools should be identified and addressed to allow successful implementation of mitigation measures by schools during future outbreaks.
Schools that reported substantial spring 2009 ILI were more likely to also report implementing the two key recommendations about planning for pH1N1. Many of these schools may have implemented the two key recommendations on planning before their schools experienced any significant pH1N1 disease; however some of these schools may have implemented the two recommendations in reaction to pH1N1 after experiencing outbreaks during the spring 2009. Guidance is needed to effectively integrate preparedness into everyday activities of schools to improve school responses during future influenza outbreaks.
Although fewer than half of respondents reported a substantial amount of illness in their school during the 2009 spring term, the majority of respondents reported that their interventions during the 2009 fall term made a difference in preventing influenza in their schools. This perception might have been influenced by the fact that there was little disease from pH1N1 in NYC public schools during fall term 2009 compared with the spring term 2009 
. The majority of respondents perceived NPIs as being effective in preventing influenza transmission. This perception could be due in part to the effort made by local and national public health authorities to promote school mitigation measures. This point is also highlighted by the fact that faculty and staff used their own funds to purchase hand sanitizer and tissue for their students in more than half of schools in the survey.
Schools reported communicating with students and parents using different methods and languages, but it is unknown how many students, parents or guardians received the communications from schools. In NYC, schools reported using mostly traditional methods of communication, including school wide handouts, letters, and student assemblies. Electronic communication methods such as e-mailing, mass texting, World Wide Web, social media were seldom used. Expanding the use of electronic methods of information sharing may enhance communication with students and parents during future influenza outbreaks. CDC is currently conducting a study to evaluate communication between schools and parents during the pH1N1 outbreak in Michigan.
This online survey had a number of limitations. Although the voluntary survey was e-mailed to all NYC public school principals by school officials, less than one-quarter of schools accessed the online survey, nearly all of whom responded to the survey. This suggests that the online format of administering the survey may have impacted the rate of participation by schools. In addition, there were only 2 weeks available to administer the survey and it was not possible to determine if each school received the study information sent via an email by NYCDOE using a general electronic mailing list. In similar studies done in the states of Georgia, Pennsylvania and Michigan, where a combination of web- and paper-based surveys were used, the responses rates ranged from 35% to 44% 
. In NYC, the response rate varied by the grade levels of the school and the school size. For example, schools with lower grade levels (K-5) were more likely to respond than schools with higher grade levels (9–12 grades). This variation could be due in part to the fact that young children, who would typically be in grades K-5, were disproportionately affected by pH1N1 during spring 2009. The survey was completed by one or a few people familiar with administrative and health services at the school, but responses may not accurately reflect the plans, actions, and experiences of all school officials. Because answering some questions in the survey required recalling information, this evaluation may be subject to recall and social-desirability bias. Additionally, respondents did not provide data for all survey questions. Moreover, most respondents were from Queens, a borough that experienced substantial pH1N1 activity in the spring of 2009 and hence, the findings from our sample may not be generalizable to the entire public school system in NYC. However, because the school system in NYC is fairly centralized and the resources needed to implement NPIs were provided by the NYC school and public health officials, it is less likely that the findings from this study would have been significantly different if more schools from other boroughs participated in the survey. Measuring the effects of individual strategies used by OSH to disseminate pH1N1 mitigation guidelines to schools is beyond the scope of this evaluation and should be assessed to help the OSH and the school system determine the most effective strategies for disseminating future mitigation guidelines. Finally, we were unable to assess the impact of the recommended NPIs on pH1N1 transmission and ILI.