TY - JOUR AU - Birk-Urovitz, Elizabeth AU - Li, Ye AU - Drews, Steven AU - Sikora, Christopher AU - Hinshaw, Deena AU - Biel, Rita K. AU - Habib, Faiza AU - Riviera, Laura AU - Usman, Hussain AU - Strong, David AU - Johnson, Ian PY - 2017/05/02 Y2 - 2024/03/29 TI - Correlation of school absenteeism and laboratory results for Flu A in Alberta, Canada JF - Online Journal of Public Health Informatics JA - OJPHI VL - 9 IS - 1 SE - Communicable Disease Surveillance Use Cases for Human, Animal, and Zoonotic Diseases DO - 10.5210/ojphi.v9i1.7679 UR - https://ojphi.org/ojs/index.php/ojphi/article/view/7679 SP - AB - <div style="left: 72px; top: 312.606px; font-size: 11.3333px; font-family: sans-serif; transform: scaleX(1.08137);" data-canvas-width="51.022666666666666">Objective</div><div style="left: 84px; top: 324.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.985455);" data-canvas-width="304.3906666666669">To assess the correlations between weekly rates of elementary</div><div style="left: 72px; top: 337.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.968664);" data-canvas-width="315.962">school absenteeism due to illness (SAi) and percent positivity for</div><div style="left: 72px; top: 351.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.03688);" data-canvas-width="315.401">influenza A from laboratory testing (PPFluA) when conducted at a</div><div style="left: 72px; top: 364.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.00231);" data-canvas-width="272.26066666666657">city level from September to December over multiple years.</div><div style="left: 72px; top: 389.94px; font-size: 11.3333px; font-family: sans-serif; transform: scaleX(1.11794);" data-canvas-width="66.10733333333333">Introduction</div><div style="left: 84px; top: 401.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.01002);" data-canvas-width="303.0737333333333">Rates of student absenteeism in schools have been mainly used to</div><div style="left: 72px; top: 415.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.03475);" data-canvas-width="315.66733333333326">detect outbreaks in schools and prompt public health action to stop</div><div style="left: 72px; top: 428.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.994396);" data-canvas-width="84.29166666666669">local transmission</div><div style="left: 156.289px; top: 428.867px; font-size: 6.8px; font-family: serif; transform: scaleX(1.00196);" data-canvas-width="8.5">1,2</div><div style="left: 164.789px; top: 428.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.982973);" data-canvas-width="113.73000000000002">. A report by Mogto et al.</div><div style="left: 278.519px; top: 428.867px; font-size: 6.8px; font-family: serif; transform: scaleX(0.958667);" data-canvas-width="4.889199999999999">3</div><div style="left: 283.409px; top: 428.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.988854);" data-canvas-width="103.088">stated that aggregated</div><div style="left: 72px; top: 441.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.989756);" data-canvas-width="314.4784666666667">counts of school absenteeism (SAi) were correlated with PPFluA, but</div><div style="left: 72px; top: 455.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.0285);" data-canvas-width="315.45766666666657">the sample may have been biased. The purpose of this study was to</div><div style="left: 72px; top: 468.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.02958);" data-canvas-width="315.70360000000005">assess the correlation between aggregated rates of SAi and PPFluA</div><div style="left: 72px; top: 481.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.01014);" data-canvas-width="315.0779999999998">for two cities, Calgary and Edmonton, in Alberta. In such situations,</div><div style="left: 72px; top: 495.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.01606);" data-canvas-width="315.21513333333337">SAi could potentially be used as a proxy for PPFluA when there are</div><div style="left: 72px; top: 508.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.00201);" data-canvas-width="232.6166666666666">not enough samples for stable laboratory estimates.</div><div style="left: 72px; top: 533.94px; font-size: 11.3333px; font-family: sans-serif; transform: scaleX(1.07312);" data-canvas-width="46.59133333333333">Methods</div><div style="left: 84px; top: 545.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.973746);" data-canvas-width="303.72766666666666">The Alberta Real-Time Syndromic Surveillance Net (ARTSSN)</div><div style="left: 72px; top: 559.534px; font-size: 6.8px; font-family: serif;">4</div><div style="left: 75.432px; top: 559.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.04201);" data-canvas-width="312.25599999999986">collects elementary SA data from the two major school boards in</div><div style="left: 72px; top: 572.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.04244);" data-canvas-width="315.8418666666666">two cities in Alberta with populations &gt;800,000. Since reasons for</div><div style="left: 72px; top: 585.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.00559);" data-canvas-width="314.9329333333334">SA are stated, rates of SAi can be calculated. Data were obtained for</div><div style="left: 72px; top: 599.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.03142);" data-canvas-width="315.4780666666666">three years, 2012 to 2014, for each city. Laboratory data on tests of</div><div style="left: 72px; top: 612.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.01504);" data-canvas-width="315.2887999999999">respiratory agents using a standardized protocol were obtained from</div><div style="left: 72px; top: 625.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.02016);" data-canvas-width="315.33299999999997">Alberta’s Provincial Laboratory for Public Health for the same time</div><div style="left: 72px; top: 639.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.03161);" data-canvas-width="315.6559999999998">period and locations. The dates of the specimens being received by</div><div style="left: 72px; top: 652.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.04266);" data-canvas-width="315.6990666666668">the laboratory were used in this analysis. For each data source, the</div><div style="left: 72px; top: 665.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.990477);" data-canvas-width="316.5399999999999">relative proportions (SAi and PPFluA) were calculated. Data for</div><div style="left: 72px; top: 679.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.0353);" data-canvas-width="315.5233999999998">the first week of school in September and for the last two weeks of</div><div style="left: 72px; top: 692.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.950145);" data-canvas-width="315.86">December were removed for each year due to the SAi rates being</div><div style="left: 72px; top: 705.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.962331);" data-canvas-width="315.6446666666666">unstable. Linear regression models were constructed, with rates of</div><div style="left: 72px; top: 719.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.952875);" data-canvas-width="315.9574666666665">SAi predicted by PPFluA. Separate models were run for each city</div><div style="left: 72px; top: 732.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.01821);" data-canvas-width="315.2196666666666">and for each year, resulting in a total of 6 models. Percent positivity</div><div style="left: 72px; top: 745.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.00838);" data-canvas-width="314.9352">for entero-rhinoviruses (PPERV) was added to see if it improved the</div><div style="left: 72px; top: 759.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.00117);" data-canvas-width="314.8173333333333">model. The regression models were created using Excel and checked</div><div style="left: 72px; top: 772.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.967325);" data-canvas-width="316.12179999999995">in the statistical programs, SAS and R. An analysis to assess the</div><div style="left: 72px; top: 785.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.00482);" data-canvas-width="213.08933333333334">influence of a lag period was assessed using R.</div><div style="left: 72px; top: 811.273px; font-size: 11.3333px; font-family: sans-serif; transform: scaleX(1.08488);" data-canvas-width="40.93600000000001">Results</div><div style="left: 84px; top: 823.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.04135);" data-canvas-width="303.48400000000026">For each city, the provincial lab tested between 4,000 and 6,000</div><div style="left: 72px; top: 836.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.955682);" data-canvas-width="315.91666666666686">specimens each fall and SAi rates were based on denominators of</div><div style="left: 72px; top: 849.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.977171);" data-canvas-width="210.4826666666667">between 20,000 and 36,000 children. The R</div><div style="left: 282.452px; top: 850.2px; font-size: 6.8px; font-family: serif;">2</div><div style="left: 285.914px; top: 849.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.96331);" data-canvas-width="102.31960000000001">, betas, and p-values</div><div style="left: 72px; top: 863.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.974693);" data-canvas-width="316.25553333333346">for all 6 regression models are shown in Table 1. The minimum</div><div style="left: 72px; top: 876.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.01021);" data-canvas-width="317.17353333333335">correlation value was 0.693 and the maximum was 0.935. Due</div><div style="left: 72px; top: 889.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.986491);" data-canvas-width="316.30199999999996">to the strong negative correlations between PPERV and PPFluA,</div><div style="left: 72px; top: 903.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.03431);" data-canvas-width="315.6367333333332">PPERV was not retained in the models. Looking at the lag periods,</div><div style="left: 72px; top: 916.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.02949);" data-canvas-width="315.55513333333334">the maximum correlations occurred at a zero week lag in two years</div><div style="left: 72px; top: 929.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.02018);" data-canvas-width="315.2026666666667">(2012 and 2014) and at a -1 week lag in 2013. The two years with a</div><div style="left: 72px; top: 943.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.0252);" data-canvas-width="315.4021333333333">zero lag were both dominated by a H3N2 strain while the year with</div><div style="left: 72px; top: 956.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(0.978739);" data-canvas-width="314.4523999999999">mainly a H1N1 strain showed a lag of -1. Only one year of H1N1 data</div><div style="left: 72px; top: 969.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.00196);" data-canvas-width="119.60066666666667">was available for analysis.</div><div style="left: 408px; top: 312.606px; font-size: 11.3333px; font-family: sans-serif; transform: scaleX(1.10259);" data-canvas-width="68.01133333333334">Conclusions</div><div style="left: 420px; top: 324.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.00518);" data-canvas-width="304.7873333333335">We observed strong correlations between the weekly rates of</div><div style="left: 408px; top: 337.901px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.03003);" data-canvas-width="315.4633333333331">elementary SAi and PPFluA at the city level over three years, from</div><div style="left: 408px; top: 351.234px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.02322);" data-canvas-width="315.37379999999996">September to December. The reasons for the difference in lag times</div><div style="left: 408px; top: 364.567px; font-size: 11.3333px; font-family: serif; transform: scaleX(1.00166);" data-canvas-width="277.2926666666665">between the H1N1 and H3N2 seasons are being investigated.</div> ER -