Limits...
Modeling workplace contact networks: The effects of organizational structure, architecture, and reporting errors on epidemic predictions.

Potter GE, Smieszek T, Sailer K - Netw Sci (Camb Univ Press) (2015)

Bottom Line: We find that contact probability decreases with distance, and that research group membership, role, and shared projects are strongly predictive of contact patterns.Adjusting for reporting error changed the estimate of the duration distribution, but did not change the estimates of covariate effects and had little effect on epidemic predictions.Our epidemic simulation study indicates that inclusion of network structure based on architectural and organizational structure data can improve the accuracy of epidemic forecasting models.

View Article: PubMed Central - PubMed

Affiliation: California Polytechnic State University, San Luis Obispo, CA, USA; Center for Statistics and Quantitative Infectious Disease, Fred Hutchinson Cancer Research Center, Seattle, WA, USA.

ABSTRACT

Face-to-face social contacts are potentially important transmission routes for acute respiratory infections, and understanding the contact network can improve our ability to predict, contain, and control epidemics. Although workplaces are important settings for infectious disease transmission, few studies have collected workplace contact data and estimated workplace contact networks. We use contact diaries, architectural distance measures, and institutional structures to estimate social contact networks within a Swiss research institute. Some contact reports were inconsistent, indicating reporting errors. We adjust for this with a latent variable model, jointly estimating the true (unobserved) network of contacts and duration-specific reporting probabilities. We find that contact probability decreases with distance, and that research group membership, role, and shared projects are strongly predictive of contact patterns. Estimated reporting probabilities were low only for 0-5 min contacts. Adjusting for reporting error changed the estimate of the duration distribution, but did not change the estimates of covariate effects and had little effect on epidemic predictions. Our epidemic simulation study indicates that inclusion of network structure based on architectural and organizational structure data can improve the accuracy of epidemic forecasting models.

No MeSH data available.


Related in: MedlinePlus

The observed distribution of contact durations, with negative binomial and zero-inflated negative binomial fitted models. Contacts with zero duration (i.e. non-contacts) comprised 92% of all dyads and are omitted from the graph. (color online)
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4663701&req=5

Figure 2: The observed distribution of contact durations, with negative binomial and zero-inflated negative binomial fitted models. Contacts with zero duration (i.e. non-contacts) comprised 92% of all dyads and are omitted from the graph. (color online)

Mentions: In order to do this, we need to create a model for the probability distribution of the duration matrix and derive the expression for its likelihood. Then, we will express the joint likelihood of the true duration matrix and the reported duration matrix as in previous sections, and maximize the log likelihood function with trust. Our reported durations have a large number of zeroes and are overdispersed. The mean of the nonzero duration reports is 26 min, and variance 987. We could use a generalized linear model to estimate the mean of the duration distribution as a function of covariate values (McCullagh & Nelder, 1989). For this approach, we considered a negative binomial distribution and a zero-inflated negative binomial distribution (fits shown in Figure 2). Our actual distribution of contact durations has spikes at 30, 60, and 90 min, either because people tended to round their durations to these values or because based on common meeting lengths, these values are actually more frequent. The parametric forms we considered did not capture this phenomenon, so we instead categorized duration in order to avoid imposing assumptions on the duration distribution.


Modeling workplace contact networks: The effects of organizational structure, architecture, and reporting errors on epidemic predictions.

Potter GE, Smieszek T, Sailer K - Netw Sci (Camb Univ Press) (2015)

The observed distribution of contact durations, with negative binomial and zero-inflated negative binomial fitted models. Contacts with zero duration (i.e. non-contacts) comprised 92% of all dyads and are omitted from the graph. (color online)
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4663701&req=5

Figure 2: The observed distribution of contact durations, with negative binomial and zero-inflated negative binomial fitted models. Contacts with zero duration (i.e. non-contacts) comprised 92% of all dyads and are omitted from the graph. (color online)
Mentions: In order to do this, we need to create a model for the probability distribution of the duration matrix and derive the expression for its likelihood. Then, we will express the joint likelihood of the true duration matrix and the reported duration matrix as in previous sections, and maximize the log likelihood function with trust. Our reported durations have a large number of zeroes and are overdispersed. The mean of the nonzero duration reports is 26 min, and variance 987. We could use a generalized linear model to estimate the mean of the duration distribution as a function of covariate values (McCullagh & Nelder, 1989). For this approach, we considered a negative binomial distribution and a zero-inflated negative binomial distribution (fits shown in Figure 2). Our actual distribution of contact durations has spikes at 30, 60, and 90 min, either because people tended to round their durations to these values or because based on common meeting lengths, these values are actually more frequent. The parametric forms we considered did not capture this phenomenon, so we instead categorized duration in order to avoid imposing assumptions on the duration distribution.

Bottom Line: We find that contact probability decreases with distance, and that research group membership, role, and shared projects are strongly predictive of contact patterns.Adjusting for reporting error changed the estimate of the duration distribution, but did not change the estimates of covariate effects and had little effect on epidemic predictions.Our epidemic simulation study indicates that inclusion of network structure based on architectural and organizational structure data can improve the accuracy of epidemic forecasting models.

View Article: PubMed Central - PubMed

Affiliation: California Polytechnic State University, San Luis Obispo, CA, USA; Center for Statistics and Quantitative Infectious Disease, Fred Hutchinson Cancer Research Center, Seattle, WA, USA.

ABSTRACT

Face-to-face social contacts are potentially important transmission routes for acute respiratory infections, and understanding the contact network can improve our ability to predict, contain, and control epidemics. Although workplaces are important settings for infectious disease transmission, few studies have collected workplace contact data and estimated workplace contact networks. We use contact diaries, architectural distance measures, and institutional structures to estimate social contact networks within a Swiss research institute. Some contact reports were inconsistent, indicating reporting errors. We adjust for this with a latent variable model, jointly estimating the true (unobserved) network of contacts and duration-specific reporting probabilities. We find that contact probability decreases with distance, and that research group membership, role, and shared projects are strongly predictive of contact patterns. Estimated reporting probabilities were low only for 0-5 min contacts. Adjusting for reporting error changed the estimate of the duration distribution, but did not change the estimates of covariate effects and had little effect on epidemic predictions. Our epidemic simulation study indicates that inclusion of network structure based on architectural and organizational structure data can improve the accuracy of epidemic forecasting models.

No MeSH data available.


Related in: MedlinePlus