Satisfaction in Delivering Maternal and Neonatal Health Services During a Pandemic: Recommendations for Infection Preparedness and Response Protocols and Employee Support for Community Health Workers

Aim: Community health workers (CHWs) have emerged as salient health team members in disadvantaged primary settings to provide critical services to disadvantaged mothers and their newborns. There is need for empirical evidence to understand how CHWs may be supported in delivering maternal and neonatal health services during pandemics. Subject and Methods: In this study we used bivariate regression to identify the lower odds for CHW perceived satisfaction for maternal and neonatal health services, with respect to client socio-demographic characteristics, coronavirus preparedness, coronavirus responsiveness, and employee satisfaction. In addition, we used structural equation modeling to investigate if coronavirus responsiveness and employee satisfaction as mediating variables inuence the relationship between coronavirus preparedness and maternal and neonatal health services. Results: From a sample of 350 CHWs across 35 districts of Punjab, we found 30 predictors with respect to coronavirus preparedness, coronavirus responsiveness and employee satisfaction which contribute to lower odds of maternal and neonatal health services. We also identied that employee satisfaction is a key mediator in the relationship between coronavirus preparedness and maternal and neonatal health services. Conclusion: We conclude with 4 critical recommendations to support CMWs in delivering optimal services, comprising: education and training, operational support, public acceptance, and employee support and benets.


Introduction
Community health workers (CHWs) are instrumental in improving maternal and neonatal health indicators in conservative countries like Pakistan where the mobility of women is restricted due to cultural and religious interpretations and women are dependent on health services at the doorstep (Gilmore & McAuliffe, 2013). The additional and essential role that CMWs are expected to perform during the coronavirus pandemic is creating awareness and ensuring prevention for infection control in underprivileged and semi-literate communities (Bhaumik, Moola, Tyagi, Nambiar, & Kakoti, 2020). Recent scholarship highlights that there is greater vulnerability to mothers and newborns during pandemics in developing countries, and also greater risk of still-births (McClure et al., 2020).
Doorstep services in the community during pandemics are not just essential for reproductive health services, but also for protection of mother, newborn and entire families through guidance about both infection control and coronavirus symptom management (Webber & Chirangi, 2020). In Pakistan, the women CHWs play a critical role in facilitating access to primary healthcare for women, and in addition, is the only healthcare support for majority impoverished women in the country (Shaikh & Hatcher, 2005).
The Pakistan Ministry of National Health Services, Regulation and Coordination launched the National Programme for Family Planning and Primary Health Care commonly referred to as the Lady Health Workers program in 1994 which has successfully deployed over 110,000 CMWs across disadvantaged communities across Pakistan (Farooq & Arif, 2014).
After recruitment CHWs receive 15 months of training and are designated to visit 1,500 women in the community to provide antenatal, natal, and postnatal services (Hafeez, Mohamud, Shiekh, Shah, & Jooma, 2011). They are also responsible for referral to nearby health facilities and provision of health education, including infection control and prevention (Douthwaite & Ward, 2005). However, CMWs in Pakistan are known to face considerable challenges while delivering services, the main two being low pay and community resistance in accepting services from non-traditional agents (Closser & Jooma, 2013).
Another issue is the low quality training they receive at induction, and the drop in their knowledge and skill set due to non-existence of a regular training system (Oxford Policy Management, 2002).
Despite the challenges they face, evidence shows that CHWs have been effective in some indicators for maternal and neonatal health, such as: increasing tetanus coverage, immunization, attended deliveries, and exclusive breastfeeding (Jalal, 2011). With regard to areas related to infection control, CHWs have also been evidenced to improve women's awareness and practices with regard to sterilizing drinking water and improving hygiene . In the age of coronavirus, the impact of CHWs services assumes greater signi cance with regard to educating mothers in the community about corona prevention and management.
After the 18th constitutional amendment in 2011 the subject of health was devolved to provinces. In However, the major limitation is that no formal training has carried out, and only guidelines for prevention, sanitation and symptom management have been distributed through booklets to CHWs. There has been no investigation about the e cacy and limitations of these guidelines, additional needs for preparedness and response, or the quality of primary healthcare services for maternal and newborn health during the pandemic.

Study aims
The perceived satisfaction of CHWs in delivering maternal and neonatal health services is an important indicator of mother and newborn wellbeing (Wilford et al., 2018). In the absence of su cient research during the coronavirus pandemic (Singhal, 2020), it is important for independent researchers to help in lling the gaps about how maternal and neonatal health services are in uenced by coronavirus preparedness and responsiveness in community health workers. At rst step we aimed to identify the lower odds for CHW perceived satisfaction for maternal and neonatal health services, with respect to four areas: 1. socio-demographic characteristics; 2. coronavirus preparedness; 3. coronavirus responsiveness; and 4. employee satisfaction.
Local literature highlights that CHWs face considerable occupational challenges (Hafeez et al., 2011), and that their service response and employee satisfaction may in uence maternal and neonatal health services. Thus, at second step we also aimed to investigate the interplay among study variables by performing structural equational modeling (SEM). Thus, the objective of our fth and last research question was to examine: 5. how coronavirus responsiveness and employee satisfaction as mediating variables in uence the relationship between coronavirus preparedness (independent variable) and maternal and neonatal health services (dependent variable). We believe our study is important for not just healthcare practitioners with weak bargaining power and inadequate governance voice in developing countries (Iacobucci, 2020), but also for the poor women dependent on primary healthcare services in low-income communities of the world (Hick & Biddinger, 2020).

Methods
This study adopted a cross-sectional quantitative design. Ethics approval for this study was taken from the Institutional Review Board, Forman Christian College University. A cover letter was provided to CHWs describing the study and informed consent was taken (Panter & Sterba, 2011). No names were taken from the respondents and there was no risk to their safety. Respondents were assured that they could withdraw from the study at any point during the telephonic interview. No incentives were offered for participation in this study.

Sample
The selection criterion for this study was all currently working, government employed, CMWs called Lady Health Workers (LHWs) providing outreach services at the doorstep of the community. A total of 44,700 LHWs are deployed in the rural and urban slums of Punjab (Kayani, Khalid, & Kanwal, 2016). The target sample for this study based on Taros sampling formulae and population of LHWs in Punjab was estimated at 327 (Yamane, 1967).

Measures
The survey included questions from three standardized tools and consisted of 58 items. There were 6 questions addressing socio-demographic characteristics of respondents (Appendix A).

Coronavirus Preparedness and Responsiveness
Coronavirus preparedness and responsiveness was measured using 27 questions from the Zika Outbreak Emergency Preparedness and Response Survey (Rajiah et al., 2019). This survey includes items from a checklist developed by The Center For Disease Control and Prevention and World Health Organisation, which assesses how prepared healthcare professionals are for an pandemic outbreak. Minor modi cations for relevancy to coronavirus were made. A ve point Likert scale was used ranging from 'strongly disagree' to 'strongly agree'. A sample item for measuring coronavirus preparedness was "I know all the information about coronavirus preparedness related to my community needs", and a sample item for coronavirus responsiveness was "I can manage the common symptoms and reactions of coronavirus".

Employee Satisfaction
Employee satisfaction was measured using 13 questions from the Community Health Worker Employer Survey (Chaidez, Palmer-Wackerly, & Trout, 2018). A ve point Likert scale was used ranging from 'strongly disagree' to 'strongly agree'. This measure took into consideration employee support from coworkers and supervisors, and satisfaction with regard to workload, pay, and contractual bene ts. Items included statements like "My supervisor/team leader treats me with respect" and "My workload is reasonable". CHW satisfaction with maternal and neonatal health services delivered CHW satisfaction with maternal and neonatal health services delivered was measured using the scale Self-reported Performance of MCH Workers-Nepal (Chhetry, Clapham, & Basnett, 2005). The measure included 12 items related to satisfaction with antenatal care, postnatal care, emergency care, birthing care, and newborn care. A ve point Likert scale was used ranging from 'strongly disagree' to 'strongly agree'. The measure included items like "I am satis ed with delivery of services for prior referrals for birth care" and "I am satis ed with delivery of services for birth complications managed and/or referred".

Data Collection
We requested a list of mobile contact numbers of CHWs from the IRMNCH & NP, Punjab, and were able to gain access to a list of 1,000 numbers. The authors of the study recruited and trained 12 women research assistants for the data collection during a two week period through zoom video sessions. The research assistants were University students of Psychology experienced in data collection. The data was collected during the months of May 2020 to June 2020, using telephonic survey method, to observe physical distancing safety during coronavirus pandemic. Initially, we text messaged the entire contact list informing CHWs of research objectives, and seeking their permission for participation in the study (Delice & Practice, 2010). We followed-up with one text message when we did not receive a reply. A total of 373 CHWs replied and gave consent to be interviewed, and we were nally able to collect complete data from 350 women, making the nal response rate for this study 35%. The responding CHWs belonged to 35 of the 36 Districts of Punjab, divided into North and South (Appendix B).

Data Analysis
We used SPSS 21.0 for analysis of descriptive statistics and bivariate regression (Bryman & Cramer, 2005). The independent variables for the study include: 'coronavirus preparedness', 'coronavirus responsiveness', and 'employee satisfaction'; and the dependent variable for the study was: 'satisfaction with maternal and neonatal health services'. Reliability statistics for the scales in the study show good reliability above values of 0.71 (Terwee et al., 2007). The overall internal consistency ranged from 0.764 to 0.878 (Table 1). At rst step descriptive statistics were derived. Then study variables were compounded to assess association between variables and linear regression was calculated in order to ascertain the direction of relationship.
Next, we calculated bivariate odds regression, by recoding study variables into binary categories. We created dummy variables with '0' representing low odds of satisfaction and '1' representing higher odds of satisfaction. Signi cance of the main effects was estimated by computing the con dence levels. Pvalues of less than 0.05 were considered signi cant for this study. For adjusted odds ratios, age and serving years, as continuous variables, were held constant. In the third phase of our analysis, the complex relationships among variables, as well as their determinants were calculated along with the parameter estimates of the structural model using a path diagram. The authors used AMOS software (version 17.0) for SEM analysis (Byrne, 2001), and entered coronavirus preparedness as the independent variable and satisfaction for maternal and neonatal health services as the dependent variable. Coronavirus responsiveness and employee satisfaction were entered at as mediating variables. We opted for maximum likelihood estimation method and performed the bootstrapping on 2000 samples with 90 percent con dence intervals.

Socio-demographic regression results
In Table 4, we present the results for higher odds of CHWs satisfaction for maternal health services and neonatal health services with respect to socio-demographic characteristics. We found no signi cant associations.  do not feel con dent as a manager or coordinator of a community exposed to coronavirus (AOR: 3.14; 95% CI 1.44-6.83), (viii) are not provided opportunities to participate in peer evaluation of skills and governance planning on coronavirus (AOR: 3.50; 95% CI 1.49-8.23), (ix) are not accepted as a legitimate authority for coronavirus awareness/ prevention in the community (AOR: 3.43; 95% CI 1.91-6.15).

Structural Equation Model
Our SEM results for model t, presented in Table 8, show that all t indices are within the acceptable limit: [GFI = .998; AGFI = .977; CFI = .999; TLI = .993; RMSEA = .042]. Root mean square error of approximation (RMSEA) .042 ≤ .08 Results presented in Table 9 show that coronavirus preparedness has a direct effect on maternal health satisfaction (β = .242, p < .001) and an indirect effect on maternal health satisfaction (β = .242, p < .

Discussion
Our study variables show linear association, providing evidence that when coronavirus preparedness, coronavirus responsiveness and employee satisfaction are high, CHWs perceive their delivery of maternal and neonatal services to be better. Bivariate regression results addressing our four research questions imply considerable policy improvement is needed for CHWs with regard to preparedness and response for infection control and general employee support. Our rst research question examined a relationship between socio-demographic variables and lower odds of satisfaction with maternal and neonatal services. The results of the data analysis reveal no signi cant associations.
The second research question tested the relationship between coronavirus preparedness and lower odds of satisfaction with maternal and neonatal services. We found that CHWs have lower odds of satisfaction when they lack information and education, and training programs by the government.
Speci cally, CHWs have lower satisfaction when they are not trained about how to guide mothers about physical distancing, decontamination, and disinfection procedures. There is also lower satisfaction when there is inadequate support from local o cials and lack of information about local emergency response and who to contact during emergencies. Other research con rms that maternal health indicators show improvement when the state invests in the training and skill development of primary healthcare workers (Scott et al., 2018). Coordination and collaboration with cross-sector partners in the community is vital for emergency response and effective service delivery (Ransom, Goodman, & Moulton, 2008). Information sharing about local health teams and effective communication between health teams is vital for optimal delivery of services by CHWs. We also found that lower probability of satisfaction is associated with CHWs lack of acceptance as key leaders by the community. Local research suggests that CHWs face considerable resistance in certain communities due to patriarchal and traditional forces which prefer local healers (Jafree, 2018).
The third research question examined the relationship between coronavirus responsiveness and lower odds of satisfaction with maternal and neonatal services. Findings revealed that CHWs have lower odds of satisfaction when they are not con dent about their exact role and abilities to identify patients and manage coronavirus in the community. Other research con rms that CHWs need formal training about their roles and responsibilities for coronavirus management (Ajisegiri, Odusanya, & Joshi, 2020). We also found that lower probabilities for satisfaction were linked with speci c problems related to management of symptoms, implementing social distancing, infection spread, and hygiene literacy. Unless CHWs have training for coronavirus symptom management and infection control within communities there is greater risk of disease spread in disadvantaged communities (Perry, Zulliger, & Rogers, 2014). We also found that the odds of satisfaction were low when CHWs are not provided opportunities to participate in peer evaluation of skills and governance planning on coronavirus. Other scholarship highlights that when CHWs participate in peer evaluation and governance, there is improved service quality for maternal and child health in the primary health sector (Kaplan, Dominis, Palen, & Quain, 2013).
The fourth research question examined the relationship between employee satisfaction and lower odds of satisfaction with maternal and neonatal services. Results reveal that CHWs have lower odds of satisfaction when they do not have a good working relationship with their coworkers and are unable to learn from each other. Prior research has reported that coworkers can assist in supporting CHWs in dealing suitably with local issues and improving services (Sharma, Webster, & Bhattacharyya, 2014).
Satisfaction is also low when supervisors are disrespectful, communication is inadequate, skill development and leadership is not supported, and work-family balance is not reinforced. Other scholarship con rms that supervisor support is essential for CHW service quality, employee outcomes, and stability in family and work equation (Jaskiewicz & Tulenko, 2012). Lastly, we found that the odds of satisfaction are low when there is job insecurity and contract is inadequate. Local research shows that there is need for CMWs contracts to be improved with regard to matching income with in ation and improving their career path and professional advancement (Haq, Iqbal, & Rahman, 2008

Concluding Recommendations
CHWs can play a critical role in controlling infection and also protecting mothers and newborn during pandemics. Pakistan is lucky to have an existing CHW programme across the country, which is managed by The Ministry of National Health Services, Regulation and Coordination and the provincial health bodies responsible for community health services. The ndings of our study enable us to inform about the needs of CMWs in delivering optimal services for maternal and neonatal health during pandemics. Our recommendations are bene cial for other developing regions planning community health services for maternal and neonatal health for disadvantaged women during pandemics. We conclude with four key areas for support to improve maternal and neonatal health services by CMWs during times of pandemics (Table 10).
1. Education and Training-is critically needed to improve skill-sets for management of coronavirus in the community, identi cation of mass exposure, and development of con dence levels in CMWs as pandemic coordinators. There is also need to provide access to understandable academic material and recent scholarship related to pandemic management and infection material in CMWs who are mostly secondary school graduates and are not highly educated.
2. Operational Support-is needed to provide clear guidelines for roles and responsibilities during pandemics, and for the introduction of routine practice for pandemic management. We also recommend the formation of and regular participation in committees for emergency response and infection control and regular meetings with local o cials and community health worker teams for improved coordination and pandemic response. 3. Public Acceptance-it is essential that there is community acceptance of CMWs as legitimate authorities and key leaders for infection control for the effectiveness of service delivery. This is possible through social media and community awareness drives by established community notables like religious leaders, elected political leaders, and older and trusted male populations within the local districts.
4. Employee Support and Bene ts-we recommend consistent accountability measures of supervisors to prevent disrespect, bullying and discrimination. There is also need for increased communication and team-building initiatives with coworkers in the primary healthcare sector, such as the medical o cer in charge, lady health visitor, vaccinators, community midwives, traditional birth attendants, medical technicians and dispensers, district health o cer, and lady health supervisor. Opportunities for employee development and leadership are also needed, along with support for work-family balance in order to make service delivery more optimal. Finally, there is need for reforms with regard to job stability and career progression.