key: cord-0908646-edzujcsa
authors: Putra, Manesha; Kesavan, Ms. Malavika; Brackney, Kerri; Hackney, David N.; Roosa, Ms. Kimberlyn M.
title: Forecasting the Impact of Coronavirus Disease During Delivery Hospitalization: An Aid for Resources Utilization
date: 2020-04-25
journal: Am J Obstet Gynecol MFM
DOI: 10.1016/j.ajogmf.2020.100127
sha: f1513d58604bdec23053473983d58134eeb02d02
doc_id: 908646
cord_uid: edzujcsa

Abstract Background The ongoing Coronavirus disease (COVID-19) pandemic has severely impacted the United States. In cases of infectious disease outbreak, forecasting models are often developed for resources utilization. Pregnancy and delivery pose unique challenges, given the altered maternal immune system and the fact that the majority of American women choose to deliver in the hospital setting. Objectives The aim of our study is to forecast the incidence of COVID-19 in general population and to forecast the overall incidence, severe cases, critical cases and fatal COVID-19 cases during delivery hospitalization in the United States. Study design We use a phenomenological model with generalized logistic growth models to forecast the incidence of COVID-19 in the United States from 4/15/2020 – 12/31/2020. Incidence data from 3/1/2020 – 4/14/2020 were used to provide best-fit model solution. Subsequently, Monte-Carlo simulation was performed for each week from 3/1/2020 – 12/31/2020 to estimate the incidence of COVID-19 in delivery hospitalizations using the available data estimate. Results From 3/1/2020 – 12/31/2020, our model forecasted a total of 860,475 cases of COVID-19 in general population across the United States. The cumulative incidence for COVID-19 during delivery hospitalization is anticipated to be 16,601 (95% CI, 9,711 – 23,491) cases. Among those, 3,308 (95% CI, 1,755 – 4,861) cases are expected to be severe, 681 (95% CI, 1324 – 1,038) critical and 52 (95% CI, 23 – 81) maternal mortality. Assuming similar baseline maternal mortality rate as the year of 2018, we projected an increase in maternal mortality rate in the US to at least 18.7 (95% CI, 18.0 – 19.5) deaths per 100,000 live birth as a direct result of COVID-19. Conclusions COVID-19 infection in pregnant women is expected to severely impact obstetrical care. From 3/1/2020 – 12/31/2020, we project 3,308 severe and 681 critical cases, with about 52 COVID-19 related maternal mortalities during delivery hospitalization in the United States. These data might be helpful for counseling and resource allocation.

maternal immune system and the fact that the majority of American women choose to deliver in 23 the hospital setting. 24

Objectives: The aim of our study is to forecast the incidence of COVID-19 in general population 25 and to forecast the overall incidence, severe cases, critical cases and fatal COVID-19 cases 26 during delivery hospitalization in the United States. 27

We use a phenomenological model with generalized logistic growth models to 28 forecast the incidence of COVID-19 in the United States from 4/15/2020 -12/31/2020. 29 Incidence data from 3/1/2020 -4/14/2020 were used to provide best-fit model solution. 30

Subsequently, Monte-Carlo simulation was performed for each week from 3/1/2020 -31 12/31/2020 to estimate the incidence of COVID-19 in delivery hospitalizations using the 32 available data estimate. 33 Results: From 3/1/2020 -12/31/2020, our model forecasted a total of 860,475 cases of COVID-34 19 in general population across the United States. The cumulative incidence for COVID-19 35 during delivery hospitalization is anticipated to be 16, 601 (95% CI, 9, 491) cases. 36 Among those, 3,308 (95% CI, 1,755 -4,861) cases are expected to be severe, 681 (95% CI, 1324 37 -1,038) critical and 52 (95% CI, 23 -81) maternal mortality. Assuming similar baseline 38 maternal mortality rate as the year of 2018, we projected an increase in maternal mortality rate in 39 the US to at least 18.7 (95% CI, 18.0 -19.5) deaths per 100,000 live birth as a direct result of 40 COVID-19.

In December 2019, an outbreak of coronavirus diseases various models utilized in forecasting the incidence of infectious diseases with their benefits and 57

limitations. The phenomenological growth models use early incidence numbers to forecast future 58 incidence, which has been shown to be useful in forecasting the incidence of diseases in 59 situations with limited epidemiological data 4 . Previously, this model successfully forecasted the 60 incidence of COVID-19 in several provinces in China 5 . 61 62 Pregnant women experience unique alterations in the immune system and are often more 63 susceptible to severe respiratory infections 6 . Furthermore, the unique risks of pregnancy and 64 delivery prompt most women to deliver in the hospital setting, hence increasing exposure to 65 other hospitalized patients and healthcare workers. There are only limited data on COVID-19 66 and pregnancy, though available data do not appear to suggest increased severity of disease 67 among pregnant women 7-9 . Aside from their physiological differences, inpatient management of 68 pregnant women with respiratory diseases can be logistically challenging given the need for fetal 69 monitoring which often is not available in the intensive care unit (ICU) setting. Given these challenges, to better prepare for the peak of the pandemic, we aim to forecast the incidence of 71 COVID-19 in general population and in pregnant women within the US. 72 73

The aim of our study is to forecast the incidence of COVID-19 in US general population and to 75 forecast the overall incidence, severe cases, critical cases and fatal COVID-19 cases during 76 delivery hospitalization in the US. 77

We utilized a phenomenological model previously used to forecast the incidence of COVID-19 81 locally, in several provinces in China, and broadly, in the entire country of China 5,10 . The 82 generalized logistic growth model (GLM) extends the simple logistic growth model to 83 accommodate sub-exponential growth dynamic with an additional scaling parameter 11 . The GLM 84 is defined by the differential equation: 85 = 1 − C(t) is the cumulative cases at time t, r is the early growth rate, p is the scaling of growth 86 parameter, and K is the carrying capacity and final epidemic size. Values of p = 1 correspond to 87 exponential growth, p = 0 represents constant growth, and 0 < p < 1 defines sub-exponential 88 growth. 89 90 Incidence data of the reported confirmed cases of COVID-19 in general population within the 91 US were obtained from the Center for Disease Control and Prevention (CDC). Data from 92 3/1/2020 -4/14/2020 were used to estimate the best-fit solution using nonlinear least squares 93 fitting. Parametric bootstrap approach was used to generate uncertainty bounds around the best 94 fit solution assuming a Poisson error structure. Based on this model we generated the daily 95 COVID-19 incidence in the US until the end of 2020. 96 97 98

We reviewed available published data on COVID-19 incidence in general population and its 100 severity based on age group. We used overall data from the US, Italy and China given these are 101 the countries with the most data 12-16 . Some of these data, contain asymptomatic cases, however 102 the majority of the cases were symptomatic. Lowest and highest data from each report was used 103 as the lower and upper bound of the inputs. We incorporated the study by Breslin et al in 104

determining the ranges for pregnancy-specific severe and critical cases, though it was only 105 incorporated to 20-29-and 30-39-years age group to match the population in their study 9 . In 106 cases where the age ranges did not match our study age group, we assume equal distribution 107 among each year in an age group and estimate the proportion of our desired age range. From 108 collected literature, critical cases were defined as cases requiring critical care admission 16, 17 . 109

Severe cases were defined as requiring hospitalization in 2 studies that was derived from general 110 population 16,17 . In one series of pregnant women only however, severity was defined based on 111 clinical criteria described by Wu et al 9, 18 . 112

We identified the probability of COVID-19 in females and then based on age group incidence we 114 estimated the incidence of COVID-19 in reproductive age women (10-49 years old). Since available evidence does not suggest increased susceptibility, we assumed the incidence of 116 COVID-19 and the incidence of pregnancy for each age group are independent of each other in 117 the lowest incidence range 7, 9, 19 . With this assumption, we calculated the incidence of COVID-19 118 in pregnancy for each age group according to the multiplicative rule of probability. To account 119

for the possibility of an increased likelihood of COVID-19 during delivery, based on previous 120

reports of severe outcomes in other respiratory viruses, we modified the upper range of 121 delivering women proportion by a factor of 2.87, which was the increased odd of contracting 122 influenza in pregnancy 6,20 . Our Monte-Carlo model inputs were listed in table 1. 123

The following formulas were used to estimate the outcomes: 125 Based on the projected incidence of COVID-19 derived from the GLM model, we performed 129

Monte-Carlo simulation for each weekly cumulative incidence of COVID-19 from 3/1/2020 -130 12/31/2020. Model inputs with ranges were varied using flat distribution probability in 10,000 131 trials for each week. Based on the Monte-Carlo simulation we estimated the incidence of COVID-19, severe COVID-19 requiring extended hospitalization, critical COVID-19 and 133 fatality during delivery hospitalization. 134

We generated daily forecast for the reported incidence of COVID-19 in the US from 4/15/2020 -138 12/31/2020 based on the incidence data from 3/1/2020 -4/14/2020. Figure 1 

Based on the Monte-Carlo analysis, we estimated the weekly and cumulative incidence of 148 COVID-19 during delivery in the US (Figure 2 & 3) . During the study period, the cumulative 149 incidence was found to be 16, 601 (95% CI, 9, 491) 

From 3/1/2020 -12/31/2020, we projected 3,308 severe cases and 681 critical COVID-19 cases 172 in US women during delivery hospitalization, with about 52 maternal mortalities, during delivery 173 hospitalization. Our study predicts that the US had experienced the peak of COVID-19 cases in 174 the first 2 weeks of April both in general population as well as among delivering women. This 175 forecast for the peak of incidence was similar to data from Institute for Health Metrics and 176 Evaluation(IHME) 22 . During this period, over 50% of the expected total hospital bed and 177 resources burden of COVID-19 cases had occured. 178

Though we do not forecast state level data, we anticipate the national numbers for delivering 180 women will share a similar bicoastal distribution seen in IHME forecasting model 22 . In order to 181 further help project this data into state-level and local estimates, we have created a simple 182 calculator tool to estimate the local incidence of COVID-19 in pregnancy. Using this HTML 183 tool, a user can input the expected incidence of COVID-19 during a certain period of time and 184 the tool will estimate the number of pregnant women with total, severe and critical COVID-19 185 cases during that particular time. We also provided a rough estimate of future cases assuming 186 similar trajectory as national trend in the US. Certainly, future numbers will strongly vary 187 depending on how close a certain region trend correlates to the national trend. This tool is 188 available through bit.ly/COVID19Del. Through this link, readers will be able download the 189 HTML file and the program will run natively in your web-browser. 190

To our knowledge, we are not aware of another study to estimate the number of 193 severe, critical and fatal cases during delivery hospitalization in the US. Our prediction is based 194 on a phenomenological model to forecast the incidence of COVID-19, which is utilized when the 195 epidemiological characteristics of an outbreak is not clearly delineated as in the case of . GLM is a purely empirical model using correlations among observed data to forecast 197 statistically similar trend in the future 23 . As opposed to mechanistic models, our model did not 198 directly account for interventions (e.g. social distancing or changing in testing pattern). However, 

With the cases peaking in the beginning of April, we are bracing for the worst impact in 234 obstetric resource utilization. Though we do not have state level data, the peak incidence is 235 expected to vary by state in relation to severity and mitigation provided by each state 22 . With the 236 numbers provided in this study, we hope that we could provide and prioritize obstetrical care 237 providers with appropriate personal protective equipment (PPE). This is especially important 238 because nearly all women with COVID-19 will elect to deliver in the hospital setting, regardless 239 of the severity of disease. Thus, obstetric care providers will be facing a unique and unavoidable 240 exposure to COVID-19. Additionally, previous reports have suggested increase risks for preterm 241 birth, preeclampsia, cesarean delivery and perinatal death 24,27 . This might suggest that we will 242 also see an increase in resources utilization not directly related to COVID-19 or delivery alone. 243 244 Historical data shows that 2-4 peripartum critical care admissions occur for every 1000 245 deliveries, but during the peak of COVID-19 we estimated that an additional of 2 ICU 246 admissions per 1000 deliveries had happened. Although the peak has passed, the number of 247 critical cases will continue to linger and potentially continue to increase the strain in the medical 248 resources. Given the unique changes in physiology and the potential logistical hurdles, critical 249 care of pregnant women with COVID-19 could be challenging. Thus, obstetric care providers 250 should prepare themselves to provide insight into the multidisciplinary care of critically-ill 251 pregnant women 28 . The Society for Maternal Fetal Medicine (SMFM) has graciously provided 252 their obstetric critical care resources for free through https://education.smfm.org/. 253

In terms of maternal mortality rate, COVID-19 is projected to increase the maternal mortality 255 modestly from 17.4 deaths to 18.7 deaths per 100,000 livebirths. However, the actual maternal 256 mortality rate for the year of 2020 may be even higher than our projected number, this is because 257

it's possible that resources reallocation, reduction in face-to-face prenatal visits and COVID-19 258 economic impact could also cause an increase in maternal mortality. Most recently, United 259

Nations has demonstrated a ~30% increase in domestic violence against women worldwide 29 . 260

Finally, although currently hypothetical, the resurgence of COVID-19 remains a possibility. 261

Another outbreak this winter could increase maternal mortality rate even higher. All of these 262 factors could result in the highest maternal mortality in the modern history. 263 264

Once epidemiological parameters of COVID-19 in various circumstances are well defined, a 266 robust mechanistic model could provide a better insight into possibility and prediction of future 267 outbreaks. Our study was limited by the availability of data from pregnant women with COVID-268 19. We encourage obstetrical care providers nationally to participate in registries for pregnant 269 women affected by COVID-19. Though we were able to estimate the impact of COVID-19 on delivery hospitalization, the impact of COVID-19 could extend beyond the disease itself. 271

Multipole guidelines have suggested to scale down antenatal testing and prenatal visits 6,27 . 272 Studies are needed to assess the impact of resources reallocation to non-COVID-19 affected 273 women, including the impact of domestic violence which was known to affect pregnant women 274 at higher rates 30 . 275 276 277

In summary, COVID-19 is also expected to severely impact obstetrical care. Despite their 279 younger age, we still projected an increase in critical care admission during delivery 280 hospitalization and maternal mortality rate among pregnant women. In the period of 3/1/2020 -281 12/31/2020, we project 3,308 severe and 681 critical cases, with about 52 COVID-19 related 282 maternal mortalities during delivery hospitalization in the United States. Continuing efforts in 283 mitigating the downstream effects of COVID-19 need to be made to prevent worsening of these 284 projected numbers. 

Severe acute respiratory syndrome 287 coronavirus 2 (SARS-CoV-2) and coronavirus disease-2019 (COVID-19): The epidemic and 288 the challenges

Applying infectious disease forecasting to public 293 health: a path forward using influenza forecasting examples

Using phenomenological models for forecasting 296 the 2015 Ebola challenge

Short-term Forecasts of the COVID-19 Epidemic in 298

ISUOG Interim Guidance on 2019 novel coronavirus 301 infection during pregnancy and puerperium: information for healthcare professionals

Clinical characteristics and intrauterine vertical 304 transmission potential of COVID-19 infection in nine pregnant women: a retrospective 305 review of medical records

Clinical manifestations and outcome of SARS-CoV-2 307 infection during pregnancy

COVID-19 infection among 309 asymptomatic and symptomatic pregnant women: Two weeks of confirmed 310 presentations to an affiliated pair of New York City hospitals

Real-time forecasts of the COVID-19 epidemic in China from 313

A generalized-growth model to characterize the early 315 ascending phase of infectious disease outbreaks

Coronavirus Disease 2019 (COVID-19) in Italy. JAMA. 2020:Ahead 317 of print

the epidemiological characteristics of an outbreak of 2019 novel 319 coronavirus diseases (COVID-19)-China

Case-Fatality Rate and Characteristics of Patients Dying 321 in Relation to COVID-19 in Italy. JAMA. 2020:Ahead of print

Demographic science aids in understanding the 323 spread and fatality rates of COVID-19. medRxiv. 2020:Ahead of print, pending peer-324 review

Estimates of the severity of coronavirus disease 326 2019: a model-based analysis. The Lancet Infectious Diseases. 2020:Ahead of print

Severe Outcomes Among Patients with Coronavirus Disease

Characteristics of and Important Lessons From the Coronavirus 331 Disease 2019 (COVID-19) Outbreak in China. JAMA. 2020:Ahead of print

Clinical manifestations and outcome of SARS-CoV-2 333 infection during pregnancy

Epidemiology of influenza in pregnant 335 women hospitalized with respiratory illness in Moscow

Births: Final data for 2018. National Vital 338 Statistics Reports

Forecasting COVID-19 impact on hospital bed-days, ICU-days, 340 ventilator-days and deaths by US state in the next 4 months. medRxiv. 2020:Ahead of 341 print, pending peer-review

Bridging Mechanistic and Phenomenological Models of Complex 343 Biological Systems

Outcome of Coronavirus spectrum infections 345 (SARS, MERS, COVID 1 -19) during pregnancy: a systematic review and meta-analysis

Universal Screening for SARS-CoV-2 in Women 348 Admitted for Delivery

Projecting the transmission 350 dynamics of SARS-CoV-2 through the postpandemic period

MFM Guidance for COVID-19

Physiologic changes in pregnancy and their impact on critical 354 care

United Nations 356 Publications

Acknowledging a persistent truth: domestic violence in pregnancy