Shining the light on abortion: Drivers of online abortion searches across the United States in 2018

The (mis)use of Google Trends data in the social sciences - A systematic review, critique, and recommendations

Researchers increasingly use aggregated search data from Google Trends to study a wide range of phenomena. Although this new data source possesses some important practical and methodological benefits, it also carries substantial challenges with respect to internal validity, reliability, and generalizability. In this paper, we describe and assess the existing applied research with Google Trends data in the social sciences. We conduct a systematic literature review of 360 studies using Google Trends data to (1) illustrate habits and trends and (2) examine whether and how researchers take the identified challenges into account. The results show that the large majority of the literature fails to test the internal validity of their Google Trends measure, does not consider whether their data are reliable across samples, and does not discuss the generalizability of their results. We conclude by stating practical recommendations that will help researchers to address these issues and properly work with Google Trends data.

The impact of abortion bans on short-term housing needs

OBJECTIVES

State-level abortion bans in the United States have created a complex legal landscape that forces many prospective patients to travel long distances to access abortion care. The financial strain and logistical difficulties associated with travelling out of state for abortion care may present an insurmountable barrier to some individuals, especially to those with limited resources. Tracking the impact of these abortion bans on travel and housing is crucial for understanding abortion access and economic changes following the Dobbs U.S. Supreme Court decision.

STUDY DESIGN

This study used occupancy data from an average of 2,349,635 (standard deviation = 111,578) U.S. Airbnb listings each month from October 1st, 2020, through April 30th, 2023, to measure the impact of abortion bans on travel for abortion care and the resulting economic effects on regional economies.

METHODS

The study used a synthetic difference-in-differences design to compare monthly-level occupancy rate data from 1-bedroom entire-place Airbnb rentals within a 30-min driving distance of abortion clinics in states with and without abortion bans.

RESULTS

The study found a 1.4 percentage point decrease in occupancy rates of Airbnbs around abortion clinics in states where abortion bans were in effect, demonstrating reductions in Airbnb use in states with bans. In the 6-month period post Dobbs, this decrease translates to 16,548 fewer renters and a $1.87 million loss in revenue for 1-bedroom entire-place Airbnbs within a 30-min catchment area of abortion facilities in states with abortion restrictions.

CONCLUSION

This novel use of Airbnb data provides a unique perspective on measuring demand for abortion and healthcare services and demonstrates the value of this data stream as a tool for understanding economic impacts of health policies.

Trends of intrauterine device insertion and 'Googling' about intrauterine devices before and during the COVID-19 pandemic in Australia

Objective

The COVID-19 pandemic significantly disrupted access to primary care in Australia. This could have negatively impacted reproductive health services rates such as intrauterine device insertion rates, and interest in seeking information about intrauterine devices by searching on Google. We aimed to assess the trends of, and the association between, the actual Medicare service utilization rates for intrauterine device insertion and searching about intrauterine devices on Google, before and during the COVID-19 pandemic.

Methods

We conducted systematic analyses of secondary data from June 2017 to May 2022, using Medicare and Google Trends data sources. We visualized the rates of intrauterine device insertion, plus Google's search volumes about 'Intrauterine device' and 'Progestin IUDs' as topics. Then, we assessed the correlation between intrauterine device insertion rates and Google search, using Spearman correlation.

Results

The average yearly rates of intrauterine device insertion increased noticeably from 25.1-26.3 in 2018-2019 to 29.3-31.2 per 100,000 population in 2020-2021 (12-18% increase). The highest monthly intrauterine device insertion rate nationally (37 per 100,000 population) was seen in March 2021. By June 2020, search term use for the two intrauterine device-related topics returned to much higher levels (50% increase for 'Progestin IUDs', and 54% for 'Intrauterine device', respectively). A moderately strong correlation was seen between actual intrauterine device insertion rates and search on Google about intrauterine devices (Spearman rho = 0.61, p < 0.000).

Conclusion

We demonstrated a moderately strong correlation between trends of intrauterine device insertion rates and search on Google about intrauterine devices, before and during the COVID-19 pandemic in Australia. Googling about intrauterine devices could, therefore, be a useful indicator to gauge future interest in actual intrauterine device insertion for months thereafter.

Experiences seeking, sourcing, and using abortion pills at home in the United States through an online telemedicine service

A growing number of people in the United States seek to self-manage their abortions by self-sourcing abortion medications online. Prior research focuses on people's motivations for seeking self-management of abortion and experiences trying to obtain medications. However, little is known about the experiences of people in the U.S. who actually complete a self-managed abortion using medications they self-sourced online. We conducted anonymous in-depth interviews with 80 individuals who sought abortion medications through Aid Access, the only online telemedicine service that provides abortion medications in all 50 U.S. states. Through grounded theory analysis we identified five key themes: 1) participants viewed Aid Access as a "godsend"; 2) Fears of scams, shipping delays, and surveillance made ordering pills online a "nerve-racking" experience; 3) a "personal touch" calmed fears and fostered trust in Aid Access; 4) participants were worried about the "what ifs" of the self-managed abortion experience; and 5) overall, participants felt that online telemedicine met their important needs. Our findings demonstrate that online telemedicine provided by Aid Access not only provided a critical service, but also offered care that participants deemed legitimate and trustworthy.

The influence of advertising policy on information available online for abortion self-referral

OBJECTIVE

Prior work shows that ads related to abortion services often feature crisis pregnancy centers instead of abortion providers. We investigated whether a change in Google's advertising policy that required advertisers to disclose whether they provided abortion services increased the proportion of ads facilitating abortion self-referral.

STUDY DESIGN

We used a standardized protocol to search online for abortion services before, during, and after the policy change; we performed searches in August 2016 to June 2017, June 2019, and October 2019, respectively, using Google, Bing, and Yahoo search engines. We performed searches for the 25 most populous U.S. cities and the 43 state capitals not already included. We classified up to the first 5 ads as facilitating abortion referral, hindering abortion referral, or providing neutral content. We compared search engine results using a chi-square test.

RESULTS

Among ads returned by Google, those shown after the policy change were significantly more likely to facilitate abortion self-referral (66.7% vs. 44.2%; p = 0.003) and slightly less likely to hinder abortion self-referral (33.3% vs. 40.6%; p = 0.33) compared to before the change. These findings were reversed for ads shown by Bing and Yahoo; ads returned after the change were significantly less likely to facilitate abortion self-referral (24.6% vs. 32.8%; p = 0.01) and significantly more likely to hinder self-referral (28.3% vs. 21.6%; p = 0.03) compared to before the change.

CONCLUSION

A policy requiring advertisers to disclose whether they provide abortion services was associated with increasing the proportion of ads facilitating self-referral. Similar policies should be considered by all search engines.

Emerging Technology: Preparing Tomorrow's MCH Workforce to Innovate for Equity

Purpose

This commentary proposes a new direction to train the MCH workforce by leveraging today’s rapidly changing innovation and technology to address persistent health inequities.

Description

We outline the creation of an MCH technology and innovation training pipeline developed by harnessing creative funding opportunities, diversifying training modalities, and expanding partnerships beyond traditional academic-practice partners, that be replicated and adapted by other academic programs.

Assessment

Technology and innovation will continue to be a growing intersection between health and equity, and we must create a robust pipeline of MCH leaders prepared to collaborate with entrepreneurial and innovation leaders.

Conclusion

Technology offers an important opportunity to improve MCH outcomes and reduce disparities, but only if we train the MCH workforce to seize these opportunities.

Exploring Google Searches for Out-of-Clinic Medication Abortion in the United States During 2020: Infodemiology Approach Using Multiple Samples

Background

As access barriers to in-person abortion care increase due to legal restrictions and COVID-19-related disruptions, individuals may be turning to the internet for information and services on out-of-clinic medication abortions. Google searches allow us to explore timely population-level interest in this topic and assess its implications.

Objective

We examined the extent to which people searched for out-of-clinic medication abortions in the United States in 2020 through 3 initial search terms: home abortion, self abortion, and buy abortion pill online.

Methods

Using the Google Trends website, we estimated the relative search index (RSI)-a comparative measure of search popularity-for each initial search term and determined trends and its peak value between January 1, 2020, and January 1, 2021. RSI scores also helped to identify the 10 states where these searches were most popular. We developed a master list of top search queries for each of the initial search terms using the Google Trends application programming interface (API). We estimated the relative search volume (RSV)-the search volume of each query relative to other associated terms-for each of the top queries using the Google Health Trends API. We calculated average RSIs and RSVs from multiple samples to account for low-frequency data. Using the Custom Search API, we determined the top webpages presented to people searching for each of the initial search terms, contextualizing the information found when searching them on Google.

Results

Searches for home abortion had average RSIs that were 3 times higher than self abortion and almost 4 times higher than buy abortion pill online. Interest in home abortion peaked in November 2020, during the third pandemic wave, at a time when providers could dispense medication abortion using telemedicine and by mail. Home abortion was most frequently queried by searching for Planned Parenthood, abortion pill, and abortion clinic, presumably denoting varying degrees of clinical support. Consistently lower search popularity for self abortion and buy abortion pill online reflect less population interest in mostly or completely self-managed out-of-clinic abortions. We observed the highest interest for home abortion and self abortion in states hostile to abortion, suggesting that state restrictions encourage these online searches. Top webpages provided limited evidence-based clinical content on self-management of abortions, and several antiabortion sites presented health-related disinformation.

Conclusions

During the pandemic in the United States, there has been considerably more interest in home abortions than in minimally or nonclinically supported self-abortions. While our study was mainly descriptive, showing how infrequent abortion-related search data can be analyzed through multiple resampling, future studies should explore correlations between the keywords denoting interest in out-of-clinic abortion and abortion care measures and test models that allow for improved monitoring and surveillance of abortion concerns in our rapidly evolving policy context.

Using Application Programming Interfaces (APIs) to Access Google Data and Gain Insights Into Searches on Birth Control in Louisiana and Mississippi, 2014-2018: Infoveillance Study

Background

It is now common to search for health information online. A 2013 Pew Research Center survey found that 77% of online health seekers began their query at a search engine. The widespread use of online health information seeking also applies to women’s reproductive health. Despite online interest in birth control, not much is known about related interests and concerns reflected in the search terms in the United States.

Objective

In this study, we identify the top search terms on Google related to birth control in Louisiana and Mississippi and compare those results to the broader United States, examining how Google searches on birth control have evolved over time and identifying regional variation within states.

Methods

We accessed search data on birth control from 2014-2018 from 2 Google application programming interfaces (APIs), Google Trends and Google Health Trends. We selected Google as it is the most commonly used search engine. We focused our analysis on data from 2017 and compared with 2018 data as appropriate. To assess trends, we analyzed data from 2014 through 2018. To compare the relative search frequencies of the top queries across Louisiana, Mississippi, and the United States, we used the Google Health Trends API. Relative search volume by designated marketing area (DMA) gave us the rankings of search volume for each birth control method in each DMA as compared to one another.

Results

Results showed that when people searched for “birth control” in Louisiana and the broader United States, they were searching for information on a diverse spectrum of methods. This differs from Mississippi, where the data indicated people were mainly searching for information related to birth control pills. Across all locations, searches for birth control pills were significantly higher than any other queries related to birth control in the United States, Louisiana, and Mississippi, and this trend remained constant from 2014 to 2018. Regional level analysis showed variations in search traffic for birth control across each state.

Conclusions

The internet is a growing source of health information for many users, including information on birth control. Understanding popular Google search queries on birth control can inform in-person discussions initiated by family planning practitioners and broader birth control messaging campaigns.

International Registered Report Identifier (IRRID)

RR2-10.2196/16543

Quality of top webpages providing abortion pill information for Google searches in the USA: An evidence-based webpage quality assessment

BACKGROUND

In the United States, the internet is widely used to seek health information. Despite an estimated 18 million Google searches on abortion per year and the demonstrated importance of the abortion pill as an option for pregnancy termination, the top webpage search results for abortion pill searches, as well as the content and quality of those webpages, are not well understood.

METHODS

We used Google's Custom Search Application Programming Interface (API) to identify the top 10 webpages presented for "abortion pill" searches on August 06, 2018. We developed a comprehensive, evidence-based Family Planning Webpage Quality Assessment Tool (FPWQAT), which was used to assess webpage quality for the five top webpages presenting text-based educational content.

RESULTS

Of the top webpages for "abortion pill" searches, a plannedparenthood.com page was the top result and scored highest on our assessment (81%), providing high-quality and useable information. The other four webpages, a Wikipedia.com page and three anti-abortion information webpages, scored much lower on our assessment (14%-43%). These four webpages had lower quality of information in less useable formats. The anti-abortion pages also presented a variety of disinformation about the abortion pill.

CONCLUSIONS

Both the lack of accurate clinical content on the majority of top webpages and the concerning disinformation they contained raise concerns about the quality of online abortion pill information, while underlining challenges posed by Google search results to informed choice for consumers. Healthcare providers and consumers must be informed of online abortion pill content that is not based in current clinical evidence, while advocates and policymakers should push for online information that is credible and useable. These changes are imperative given the importance of sound abortion pill information for reproductive decision-making at a time when in-person abortion services are further challenged in the US.