Early evaluation of the ‘STOP SEPSIS!’ WHO Global Maternal Sepsis Awareness Campaign implemented for healthcare providers in 46 low, middle and high-income countries

Objective To evaluate changes in awareness of maternal sepsis among healthcare providers resulting from the WHO Global Maternal Sepsis Study (GLOSS) awareness campaign. Design Independent sample precampaign/postcampaign through online and paper-based surveys available for over 30 days before campaign roll-out (pre) and after study data collection (post). Descriptive statistics were used for campaign recognition and exposure, and odds ratio (OR) and percentage change were calculated for differences in awareness, adjusting for confounders using multivariate logistic regression. Setting and participants Healthcare providers from 398 participating facilities in 46 low, middle and high-income countries. Intervention An awareness campaign to accompany GLOSS launched 3 weeks prior to data collection and lasting the entire study period (28 November 2017 to 15 January 2018) and beyond. Main outcome measures Campaign recognition and exposure, and changes in awareness. Results A total of 2188 surveys were analysed: 1155 at baseline and 1033 at postcampaign. Most survey respondents found the campaign materials helpful (94%), that they helped increase awareness (90%) and that they helped motivate to act differently (88%). There were significant changes with regard to: not having heard of maternal sepsis (−63.4% change, pre-OR/post-OR 0.35, 95% CI 0.18 to 0.68) and perception of confidence in making the right decisions with regard to maternal sepsis identification and management (7.3% change, pre-OR/post-OR 1.44, 95% CI 1.01 to 2.06). Conclusions Awareness raising campaigns can contribute to an increase in having heard of maternal sepsis and an increase in provider perception of confidence in making correct decisions. Offering the information to make accurate and timely decisions while promoting environments that enable self-confidence and support could improve maternal sepsis identification and management.


AbstrACt
Objective To evaluate changes in awareness of maternal sepsis among healthcare providers resulting from the WHO Global Maternal Sepsis Study (GLOSS) awareness campaign. Design Independent sample precampaign/postcampaign through online and paper-based surveys available for over 30 days before campaign roll-out (pre) and after study data collection (post). Descriptive statistics were used for campaign recognition and exposure, and odds ratio (OR) and percentage change were calculated for differences in awareness, adjusting for confounders using multivariate logistic regression. setting and participants Healthcare providers from 398 participating facilities in 46 low, middle and high-income countries. Intervention An awareness campaign to accompany GLOSS launched 3 weeks prior to data collection and lasting the entire study period (28 November 2017 to 15 January 2018) and beyond. Main outcome measures Campaign recognition and exposure, and changes in awareness. results A total of 2188 surveys were analysed: 1155 at baseline and 1033 at postcampaign. Most survey respondents found the campaign materials helpful (94%), that they helped increase awareness (90%) and that they helped motivate to act differently (88%). There were significant changes with regard to: not having heard of maternal sepsis (−63.4% change, pre-OR/post-OR 0.35, 95% CI 0.18 to 0.68) and perception of confidence in making the right decisions with regard to maternal sepsis identification and management (7.3% change, pre-OR/ post-OR 1.44, 95% CI 1.01 to 2.06). Conclusions Awareness raising campaigns can contribute to an increase in having heard of maternal sepsis and an increase in provider perception of confidence in making correct decisions. Offering the information to make accurate and timely decisions while promoting environments that enable self-confidence and support could improve maternal sepsis identification and management.

IntrODuCtIOn
The global health community has recently drawn attention to the importance of sepsis and its toll on global mortality and morbidity. [1][2][3] In 2017, the World Health Assembly approved a resolution on sepsis to improve the prevention, diagnosis and management of sepsis. 4 With updates in 2017 and 2018, the Surviving Sepsis Campaign has been developing guidelines for management and recommended bundles of care for sepsis among adult populations, not specific to pregnant or recently pregnant women, since 2002. [5][6][7] strengths and limitations of this study ► This study presents the results of an evaluation of a global awareness campaign which accompanied a research study on maternal infections and sepsis. ► This evaluation was a cost-effective, feasible way in which to assess campaign effectiveness among a varied and global population of healthcare providers. ► Our precampaign/postcampaign using anonymous surveys with no control group does not allow to discern the impact of the campaign alone or matching precampaign and postcampaign responses. ► Campaign fidelity was only assessed through healthcare provider self-report at postcampaign surveys. ► This evaluation was restricted to the duration of the study follow-up period limiting understanding of long-term impact.
Infections and sepsis remain the major causes of death and disability among women during pregnancy, childbirth, postpartum and post-abortion. 8 9 To respond to this, the Global Maternal and Neonatal Sepsis Initiative was launched in 2016. 4 10 Building on the 2016 SEPSIS-3 definition, 11 the World Health Organization (WHO) led the development of a definition for maternal sepsis as 'a life-threatening condition defined as organ dysfunction resulting from infection during pregnancy, childbirth, postpartum, and post-abortion'. 12 And in 2017, WHO led the Global Maternal Sepsis Study (GLOSS) and Awareness Campaign to assess the burden of maternal infections and sepsis, to validate identification criteria for possible severe maternal infection and maternal sepsis and to raise awareness on maternal sepsis among healthcare providers working in study participating facilities. 13 Awareness raising has mostly been attempted through campaigns. These have been implemented to increase knowledge, improve attitudes or change behaviours around different health issues. [14][15][16] Specific to sepsis, the UK Sepsis Trust heads a campaign on sepsis since 2012 and the Global Sepsis Alliance leads efforts aimed at raising sepsis awareness since 2010. 17 18 However, neither of these two large campaigns have been specific to maternal sepsis and to our knowledge neither has been thoroughly evaluated to assess for impact in increasing awareness.
This evaluation looked at recognition and exposure to the GLOSS campaign materials and changes in provider awareness of maternal sepsis after campaign implementation. The latter included changes in knowledge on maternal sepsis and perception of enabling environments for identification and management of maternal sepsis.

MethODs
The GLOSS campaign was designed to accompany the Global Maternal Sepsis Study with the goal of raising awareness on maternal sepsis among healthcare providers working in participating facilities. Details regarding study protocol, including selection of countries and facilities, were published elsewhere. 13 In short, GLOSS was a facility-based, 1-week inception cohort study which enrolled pregnant or recently pregnant women with suspected or confirmed infection at 713 healthcare facilities in prespecified geographical areas located in 52 low, middle and high-income countries. 13 Color key: teal: countries included in the GLOSS STOP SEPSIS! awareness campaign evaluation (N=37); green: countries eligible for the evaluation but excluded from this analysis because less than 2 responses received (N=9) The boundaries shown on this map do not imply the expression of any opinion whatsoever on the part of WHO concerning the legal status of any country, or concerning the delimitation of its frontiers or boundaries the stOP sePsIs! awareness campaign The campaign launch was planned for before study implementation continuing throughout data collection and beyond. It was designed using existing frameworks for public information campaigns, social marketing, health communication and behaviour change. 14 19-21 The development of the campaign included an overarching communication strategy using a multicomponent approach delivering a simple and consistent message through visually attractive media. 22 The campaign had a soft launch with an online congress on 12 September 2017 and the full campaign roll-out began on 6 November 2017, which included a website, printed materials, social media messaging and press releases. While global coordination of the campaign was undertaken by WHO, implementation of the campaign was the remit of GLOSS country coordinators. Box 1 describes the different actions and components that were necessary for the design and development of the STOP SEPSIS! awareness campaign.

evaluation of the stOP sePsIs! awareness campaign
We used an independent sample precampaign/postcampaign design through online and paper-based surveys. Details regarding the definition used for awareness for this campaign, survey formulation and dissemination, including analysis of baseline data, have been published elsewhere. 23 Briefly, a precampaign 32-question survey was developed to gather baseline information on healthcare providers' awareness of maternal sepsis through self-reported knowledge on maternal sepsis and perception of their work environments as enabling for the identification and management of maternal sepsis. Knowledge was assessed through questions relating to whether respondents had heard of maternal sepsis, correct identification of criteria that define maternal sepsis (infection plus organ dysfunction) and identification of correct initial management of maternal sepsis and infections (antibiotics and fluids) when maternal sepsis was suspected in the case vignette presented in the survey. Perception of enabling environments was assessed through self-reported confidence in making right decisions, reported availability of resources for correct identification and management, and feeling of support from their work environments in dealing with maternal sepsis, using a 5-point Likert scale. The same survey was administered at postcampaign to assess changes in knowledge and perception of their environments; 14 additional questions were included in the postsurvey which considered respondents' recognition of and exposure to the campaign, such as knowledge about the study and the campaign, message recall, engagement with social media for the campaign and whether the campaign materials prompted changes in behaviour. See online supplementary appendix 1 for a copy of the surveys.
Eligible respondents were healthcare providers working in GLOSS participating facilities in countries that received financial support for campaign implementation (n=46); we excluded all surveys from respondents that did not explicitly state that they were providers caring for women with infections in healthcare facilities (eg, hospital administrators, or community health workers, or if the field was left blank) and from countries with less than two responses at either precampaign or postcampaign (n=9). See figure 1 for a map of all the countries eligible for this evaluation. The surveys were distributed using a snowballing technique and were available in seven languages: Arabic, English, French, Portuguese, Russian, Spanish and Vietnamese. The surveys were available for over 30 days (precampaign between 29 September and 5 November 2017, postcampaign between 31 January and 11 March 2018). Weekly reminders were sent through the online tool and via email to non-respondents. Targeted outreach was undertaken in countries with fewer than two responses. The campaign was active between 6 November 2017 and 15 January 2018; however, countries were encouraged to continue to use the materials beyond GLOSS study implementation.

Data analysis
We used descriptive analysis to provide frequencies and percentages for the characteristics of the sample, knowledge and perceived enabling environments and for all the questions relating to campaign recognition and exposure. The latter were assessed through postcampaign surveys only and complemented with self-reported accounts by GLOSS country coordinators. Text-based responses were codified into numerical values according to common emerging themes. All Likert scale answers were dichotomised assigning a 1 to the two most favourable responses (ie, they felt very or somewhat confident about making the right decision) and 0 to the combination of remaining options (neutral, not very confident or not confident at all). While previously we assessed dichotomisation using 1 to the single most favourable response (ie, respondent felt very confident) and a 0 to the combination of remaining options (somewhat confident, neutral, not very confident or not confident at all) 23 we decided to include a more flexible definition of confidence, perception of availability of resources and feeling of support to allow for a more robust denominator that would enable comparisons. Figure 3 Responses when answering Yes to the question 'Did the information provided in the materials motivate you to do something differently than before?' (n=658). (Respondents were able to check as many response options as needed.) See online supplementary appendix 2 for results of the overall analysis using this second dichotomisation not used in this evaluation.

Open access
To assess the impact of the campaign we conducted several analyses. First, we calculated percentage change ([(% in post − % in pre)/% in pre]×100) and estimated ORs to determine differences in respondent knowledge and perception of enabling environments after campaign implementation relative to baseline measure for the total sample and by respondent characteristics. Due to the methodology used for survey dissemination and anonymity of surveys, this was not a matched sample, paired response precampaign/postcampaign. Sensitivity analyses were conducted restricting the population to facilities from which we received at least one survey response and at least two responses, and to countries with more than 30 survey responses per country at precampaign and at postcampaign.
Second, we used multivariate logistic regression models to explore the association between respondents and facilities' characteristics at precampaign and postcampaign and change in components of awareness after campaign implementation. Based on analysis of baseline data and our assumptions on characteristics that would be associated with levels of awareness, 23 we included the following variables in the model: whether respondent was a physician, years of work experience, region where the respondent worked, whether the country had implemented an expanded version of the campaign and whether the facility was a level III facility. Countries were considered to have implemented an expanded version of the campaign if they had printed and displayed all posters and infographics, prepared and disseminated press releases and if they had organised other activities or developed other materials for the campaign. Since less than 20% of respondents participated in the World Sepsis Congress Spotlight we did not include this variable in our models. We looked at effect modification by examining interactions between the time of the survey (precampaign or postcampaign) and each of the characteristics included in the model.
We used Pearson's χ 2 test to compare proportions and Wald's test to assess for significant differences in the models including interaction terms. Logistic regression analysis was used to estimate ORs between pre-and post-, crude and adjusted, clustering at the geographical area level. Statistical significance is reported at p<0.05. Stata (V.14.2, College Station, TX) was used for the analyses.

Patient and public involvement
This research was done without patient or public involvement. While the development of the campaign was done with input from study regional and country coordinators, respondents to the surveys were not invited to comment on the study design or to contribute to the writing of this manuscript given their anonymity.  The funders of the study had no role in study design, data collection, data analysis, data interpretation or writing of this original article.

results
A total of 2188 surveys met our inclusion criteria. Of these, 1155 from 192 facilities were received at baseline and 1033 from 196 facilities at postcampaign. There were no significant differences in sociodemographic characteristics between respondents at precampaign and postcampaign surveys, except for a higher proportion of respondents working in a public facility at postcampaign and a higher proportion of respondents from countries where an expanded version of the campaign was implemented (table 1). Responses came from the same countries at precampaign and postcampaign. Because of the technique used for survey dissemination and because we did not know the total population of potentially exposed healthcare providers working in GLOSS participating facilities (provider turnover, rotation and replacements are high), we were unable to calculate a response rate. However, since the campaign was implemented equally at the geographical area level, if providers remained within the study area they would have been exposed to the campaign. Results from the sensitivity analyses showed that overall findings in the subgroups considered were consistent with the results from the complete sample (online supplementary appendix 3); for this reason, we used the entire sample for all subsequent analyses.
We first present the results relating to campaign recognition and exposure and then results relating to changes in knowledge and perception of respondents' work environments.
Campaign recognition and exposure Campaign recognition and exposure were high among most of the postcampaign survey respondents. Seventy-six per cent of respondents stated they noticed the materials in their facilities; among those, 94% reported finding the materials helpful, 90% that the materials helped increase awareness on maternal sepsis and 88% that the materials motivated them to do something differently. Only 8% of respondents had used Twitter to amplify the message of the campaign (figure 2). Among respondents who stated that the information provided in the materials motivated them to do something differently than before, 84% stated that it motivated them to suspect maternal sepsis and 78% to act fast (figure 3). Among respondents stating that the materials had not motivated them to do anything differently, 45% said it was because they already knew about maternal sepsis identification and management while 12% stated they had not seen the campaign materials.
Country coordinators shared anecdotal experiences of increased awareness in their facilities and the implementation of changes in practice and policies because of the study and the campaign. These accounts speak of a broader engagement with maternal sepsis identification and management. See box 2 for some examples.
Knowledge on maternal sepsis and perception of enabling environments At precampaign survey, 92% of respondents (1049/1144) had heard of maternal sepsis. However, only 16% (109/673) of respondents were able to correctly identify the definition criteria of maternal sepsis and 45% (114/251) identified the correct management for maternal sepsis. In addition, at precampaign, most survey respondents stated that their work environments were enabling for maternal sepsis identification and management: 78% (897/1155) stated that they felt confident of making right decisions, 79% (909/1155) that they perceived resources were available and 80% (921/1155) that they felt supported by their facilities. See table 2 for overall results.
After campaign implementation there was a significant decrease in respondents who stated not having heard of maternal sepsis (−63.4% change; OR 0.35, 95% CI 0.18 to 0.68). There was also a significant increase in perceived confidence in making right decisions with regard to After controlling for respondent and facility characteristics, being a physician, having less than 10 years of experience and working in a level III facility were associated with decreased odds of not having heard of maternal sepsis at precampaign (table 3). Respondents from facilities that had implemented an expanded version of the campaign were more likely to have heard of maternal sepsis and identify the correct management of maternal sepsis at postcampaign. Respondents with less than 10 years of experience were more likely to have heard of maternal sepsis at precampaign, but there were no differences across providers with different years of experience after the campaign.
Physicians were more likely to respond that they felt confident in making the right decisions at postcampaign, while being a physician and having more than 20 years of experience had a significant interaction with time of the survey with regard to perception of availability of resources and support from their facilities. At precampaign and postcampaign, respondents with 20 years or more of experience were more likely to perceive availability of resources for making right decisions and to feel supported by their facilities and these differences between groups were significant after the campaign (table 4). No differences in the perception of enabling environments were seen among respondents from facilities that had implemented an expanded version of the campaign.

DIsCussIOn
To the best of our knowledge, this is the first study to assess impact of an awareness campaign aimed at healthcare providers and implemented at a global level where precampaign and postcampaign data were collected in addition to measures relating to campaign recognition and exposure. Most healthcare providers stated that the campaign helped increase awareness of maternal sepsis and motivated them to do something differently, particularly to suspect maternal sepsis and act faster. Reports from the field also support this finding that exposure to the campaign increased sensitisation to maternal infections and sepsis. Moreover, most survey respondents had heard of maternal sepsis even before campaign implementation; after the campaign this increased significantly. Although most respondents perceived their enabling environments in a positive way before campaign implementation, there was an increase in respondent confidence to make the right decisions regarding maternal sepsis identification and management after campaign implementation.
The STOP SEPSIS! awareness campaign implementation was effective with regard to respondents' recognition of and exposure to the campaign; other campaign evaluations have used these measures to positively assess shortterm impact of campaigns. [24][25][26] Furthermore, consistent and repeat exposure to campaign messaging has shown to increase awareness 14 ; while exposure was only measured over the course of this evaluation period corresponding to the intended implementation period of the campaign, Countries were considered to have implemented an expanded version of the campaign if they had printed and displayed all posters and infographics, prepared and disseminated press releases and if they had organised other activities or developed other materials for the campaign. Adjusting for whether respondent was a physician, years of work experience, region, whether the country implemented an expanded version of the campaign and whether respondent worked in a level III facility, clustering at the geographical area level. Wald's test used to assess for differences in the models including interaction terms. *P<0.05.
aOR, adjusted OR. Countries were considered to have implemented an expanded version of the campaign if they had printed and displayed all posters and infographics, prepared and disseminated press releases and if they had organised other activities or developed other materials for the campaign. Adjusting for whether respondent was a physician, years of work experience, region, whether the country implemented an expanded version of the campaign and whether respondent worked in a level III facility, clustering at the geographical area level. Wald's test used to assess for differences in the models including interaction terms. *P<0.05.

Open access
aOR, adjusted OR.
Open access the fact that most respondents stated the campaign raised awareness is a promising trend in the right direction.
Overall knowledge about maternal sepsis increased from precampaign to postcampaign implementation among respondents to our survey with regard to having heard about maternal sepsis. Our finding that overall knowledge increased is supported by existing literature that suggests that campaigns can increase knowledge on a specific topic among healthcare providers 15 27 28 as well as among the general population. 24 29 30 The fact that there was a slight increase in identifying the correct management of maternal sepsis is important. Research has shown that knowing what is needed to manage maternal sepsis correctly and early management of maternal sepsis are critical to implementing any changes in providers' behaviour and improving maternal health outcomes. 6 31 32 The low number of providers able to identify the two criteria defining maternal sepsis might be more a reflection of the lack of consensus on this condition prior to 2017, rather than a shortcoming of the campaign. 12 The GLOSS awareness campaign was associated with reducing differences among groups of healthcare providers depending on their qualifications or years of experience. This speaks to the importance of including healthcare providers with different qualifications and years of experience in awareness raising efforts.
We found there were overall increases in respondent confidence in making right decisions about maternal sepsis identification and management, but no significant changes with regard to overall respondent perception of availability of necessary resources and feeling supported by their facilities. Evidence shows that confidence can affect clinical performance 33 and that high levels of confidence among healthcare providers can have a positive impact on patients' perception of experience of care. 34 However, the change in perception of availability of resources and support limited to physicians and more experienced providers raises a broader question on actions that facilities need to take to empower all healthcare workers in feeling that they have the necessary resources and feel supported to provide quality care. This is especially important if we consider that a more restrictive definition of enabling environments results in much lower overall levels of perceived confidence, perception of availability of resources and feeling of support. Perceived lack of availability of resources may also be a product of increased awareness of what is necessary to address maternal sepsis. These findings are a call to hospital administrators and policymakers to foster enabling environments and secure availability and access to life-saving resources.
Sepsis awareness is gaining traction on global agendas 4 10 17 ; this is supported by evidence from two studies looking at internet searches on sepsis, 35 36 meaning increases resulting from this campaign could be responding to natural trends or other factors. It is also possible that awareness was raised by having participated in the research study and not necessarily because of the campaign; disentangling the effect of the campaign from that of the implementation of the research study was impossible. Understanding whether any of these changes are sustained over time would provide us with further information on the lasting effects of the campaign.
Literature shows that while a campaign can help in raising awareness, it is insufficient in allowing for changes in behaviour. 14 20 While behaviour change is important in impacting population-level health, it is one of many components needed to make significant improvements; evidence from this study, similar to others, highlights the need for health system improvements such as availability of critical resources and support to improve maternal outcomes. 37 Assessing the impact that increased awareness resulting from a campaign has on behaviour change would provide us with supporting evidence that campaigns can help in improving health outcomes.
This study has some limitations. First, we used a precampaign/postcampaign methodology with no control group which does not allow to discern the impact of the campaign alone. Second, the method used to disseminate the survey and the fact that surveys were anonymous made it impossible to match responses at precampaign and postcampaign. Surveys were anonymous to encourage providers to respond and remove potential response bias. However, it is to note that characteristics of participants at precampaign and postcampaign were similar. Third, because implementation of the campaign was left up to country coordinators, campaign fidelity was only assessed through healthcare provider self-report at postcampaign surveys. Fourth, this evaluation was restricted to the duration of the study follow-up period, hence providing insight into early findings only and limiting our knowledge of lasting impact of the campaign, which was beyond the goal of this activity. However, our findings suggest that campaigns can have at least short-term effects on provider's knowledge and confidence. The positive perception of the campaign materials is encouraging. And fifth, since baseline data were collected after the soft launch of the campaign, the effect of the campaign may have been minimised because awareness had already been increased through exposure to the online congress as well as other global activities on sepsis conducted by other groups. However, we know that less than 20% of respondents participated in the congress.
Our findings have implications for both practice and research. On the one hand, there appear to be benefits to coupling large multicountry studies with awareness campaigns. A campaign targeting healthcare providers can promote their engagement with research studies being conducted, potentially improving study outcomes. There is also evidence that including an awareness campaign creates an environment prime to implementing changes to clinical practice as per research study protocol. On the other hand, there is a clear need for additional research to identify lasting effects of awareness campaigns, especially as global initiatives focus on increasing awareness on maternal health issues.
A campaign designed to raise awareness among healthcare providers working in facilities participating in a global research study was associated with an increase in having heard of maternal sepsis, as well as increased provider Open access perception of confidence in making correct decisions. Offering healthcare providers with the information to make accurate and timely decisions while promoting environments that enable self-confidence and support could improve maternal sepsis identification and management, which can ultimately have an impact on maternal health outcomes.
Contributors The evaluation plan and first draft of this original article was conceived by VB. MB provided substantial contributions to the data planning and analysis and first versions of this manuscript. The idea for the inclusion of an awareness campaign was first conceived by JPS. CLTR, JPS and EA provided additional feedback on first draft and data analysis. VB, MB, JPS, EA, AB, BF, MK, PL, MM, AN, NBO and ZPQ were instrumental to campaign design and implementation in their regions/countries and provided input to data collection tools used for this evaluation. All authors read and approved the final version. The WHO GLOSS Research Group contributed to different aspects of the study, from protocol development to acquisition of data and coordination of study in their countries. VB, MB and CLTR had full access to all the data and VB has final responsibility for the decision to submit for publication. Map disclaimer The depiction of boundaries on the map(s) in this article does not imply the expression of any opinion whatsoever on the part of BMJ (or any member of its group) concerning the legal status of any country, territory, jurisdiction or area or of its authorities. The map(s) are provided without any warranty of any kind, either express or implied.
Competing interests All authors declare support from WHO/SHR, Merck for Mothers and USAID for the submitted work. VB, MB and JPS were employed by WHO at the time of the study. No other relationships or activities could appear to have influenced the submitted work.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Patient consent for publication Not required.
ethics approval Ethical approval for the entire study (GLOSS), including the awareness campaign, was obtained from WHO's Ethics Review Committee in Geneva, Switzerland (protocol ID A65787) for a multisite, multicountry project, and from local and facility ethics committees as necessary. A legend was included in the survey stating that response to the survey implied consent to participate; respondents could recuse themselves at any point during the survey, including skipping and not answering certain questions.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available upon reasonable request. All enquiries regarding the data and analyses can be made to the corresponding author.
Open access This is an open access article distributed in accordance with the Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits others to copy, redistribute, remix, transform and build upon this work for any purpose, provided the original work is properly cited, a link to the licence is given, and indication of whether changes were made. See: https:// creativecommons. org/ licenses/ by/ 4. 0/.