The National Health Service (NHS) continues to face a serious discrimination issue. Naqvi, Razaq & Wilhelm (2017) identified that within the last 12 months, 13% of NHS staff reported experiencing discrimination from a colleague or manager, a 2% increase from 2015. Although black and minority ethnic (BME) staff have a higher likelihood of experiencing unfair treatment in the NHS (Kline, 2014; Jaques, 2013; Santry, 2009), both white and BME staff members reported discrimination in the workplace, at 6% and 14% respectively (Naqvi et al., 2017). Regardless of implementing initiatives to address the problem, such as The Race Equality Action Plan, few have yielded noteworthy results (Kline, 2014; Naqvi et al., 2017). This demoralizing state of affairs leads to the conclusion that discrimination is engrained in the culture of the NHS.
Previous research has indicated that correlations exist between discrimination and poor health outcomes in individuals (Harnois & Bastos, 2018; Mohammed & Williams, 2009; Brown et al., 2018). Experienced discrimination can affect both mental and physical health, including depressive symptoms, anxiety, psychiatric disorders, increased blood pressure, increased cortisol levels, cardiovascular disease and higher mortality (Harnois & Bastos, 2018; Williams & Mohammed, 2009). The clear link between experienced discrimination and its negative effects raises the question of how healthcare practitioners (HCPs) in the NHS are impacted by the discrimination they face.
Race-related attitudes and beliefs play a major role in disparities across healthcare services, especially in reference to HCPs bias towards patients (Penner et al., 2016; Ryn, Burgess, Malat & Griffin, 2006). In a recent literature review of 37 qualifying studies, Maina, Belton, Ginzberg, Singh and Johnson (2018) identify that an HCP’s explicit and implicit bias, paired with a patient’s perceived discrimination, interferes with successful patient care and patient-provider communication. Additionally, Dawson (2018) states that within the NHS “high work pressure, staff perceptions of unequal treatment, and discrimination against staff were all damaging for patient satisfaction.” The detrimental and seemingly reciprocal nature of discrimination in the workplace is well-defined in the healthcare field, whether one is seeking medical attention or reporting to her or his workplace. What is not yet clear is the extent to which experienced discrimination is associated with an HCPs bias, both internalised and explicit.
The goal of this study is (1) to estimate the prevalence of discrimination experienced by HCPs in the NHS and examine the distribution by sociodemographic and socio-economic characteristics, (2) to investigate whether an association exists between HCP’s experienced discrimination in the workplace and an HCP’s internalised stigma and (3) to explore experienced discrimination in the workplace and how it relates to explicit bias towards others. The following hypotheses will be tested:
H1: When analysing the distribution of experienced discrimination in the NHS, there will be a higher interaction between reported ethnicity, gender and age than other socio-economic and sociodemographic indicators.
H2: Experiencing discrimination in the workplace will positively associate with self-reported internalised bias within HCPs.
H3: A positive association will be found between experiencing discrimination in the workplace and higher explicit bias towards others regardless of an HCP’s own characteristics and disadvantaged statuses.
Methods
Participants
INPUT DEMOGRAPHIC INFO FROM ZOE. Recruitment – gatekeepers. The selection criteria for the survey invited any healthcare practitioner over the age of 18 with at least 12 months experience working or training in a healthcare setting to participate. The subjects were also required to have an NHS email account.
Study Design
To test the hypothesis, this dissertation will use existing survey data and follow a cross-sectional study design. A cross-sectional study will help estimate the prevalence of discrimination within the NHS and investigate the associations between discrimination experienced by HCPs as an independent variable and both internalized stigma and bias towards others as dependent variables. A cross-sectional design is appropriate based on the type of data collection and the fact that the study will be analyzed at one time point.
Materials and Procedure
This study will evaluate data from the Tackling Inequalities and Discrimination Experiences in Health Services (TIDES) study lead by King’s College London and funded by the Wellcome Trust. The study aims to investigate the inequalities in health services precipitated by discriminatory bias in order to create future interventions (TIDES Protocol). The TIDES survey, created by a team of researchers and academics at King’s College London, includes comparable measures drawn from the South East London Community Health Study (SELCoH), the Adult Psychiatric Morbidity Survey (APMS), the NHS staff survey and the Workforce Race Equality Standard (WRES) survey. The survey takes approximately 35 minutes and data collection was completed online through the online survey tool Qualtrics. The participants each receive a £15 GBP voucher upon completion.
Measures
Socio-demographic and socioeconomic indicators. The TIDES survey gathered socio-demographic and socioeconomic information from all survey participants. The study will consider age, gender, ethnicity, migration status, religious affiliation, sexual orientation, education, occupational group and household income for the analysis of discrimination prevalence and distribution within the sample.
Experienced discrimination. An adaption of a measures previously addressed in the SELCoH survey (Hatch et. Al, 2011) and WRES survey ( will assess the independent variable of experienced discrimination in the workplace. The variable will be determined by four questions: “In the last 12 months have you personally experienced discrimination at work from any of the following?” asked regarding (1) patients/service users, their relatives, or any other members of the public and (2) manager/team leader or other colleagues, both followed by the contingency question “On what grounds have you experienced discrimination? Please select the main reason. Please select a characteristic related to your own identity.” Responses for the first two questions are dichotomous (yes/no) and the responses for the multiple-choice contingency question are ethnic background, gender, religion, sexual orientation, disability, age, weight, education (including income or class), mental illness and other.
Internalised stigma. Three measures will describe the variable internalised stigma. The first is an abbreviated version of the Pre-Encounter Self-Hatred stage of the Cross Racial Identity Scale (Cross, 1971). The TIDES survey adaptation uses a 7-point Likert scale (from 1 = “strongly disagree” to 7 = “strongly agree”) to gauge the four statements: “I go through periods when I am down on myself because of my ethnic group membership; when I look in the mirror, sometimes I do not feel good about the ethnic/racial group I belong to; Privately I sometimes have negative feelings about being a member of my ethnic/racial group; I sometimes have negative feelings about being a member of my group.” The second measure is a short-form of the Internalized Homophobia scale (Meyer, 1995) also converted into a 7-point Likert scale (from 1 = “strongly disagree” to 7 = “strongly agree”) to measure the four statements: “I am comfortable about people finding out about my sexual orientation; it is important for me to control who knows about my sexual orientation; I feel comfortable discussing my sexual orientation in a public situation; even if I could change my sexual orientation I wouldn’t.”
Explicit bias. Three additional Likert scales will assess explicit bias relating to discrimination and prejudice in modern society. The first is a 5-point scale which asks participants to rate (from 1 = “strongly disagree” to 5 “strongly agree”) 10 positive and negative statements about the treatment of ethnic minorities in the UK. The second scale addresses feelings towards discrimination and prejudice faced by LGBT+ people on a 7-point scale (from 1 = “strongly disagree” to 7 “strongly agree”). The final Likert scale returns to a 5-point scale with 10 statements about individually held biases and feelings towards prejudice including, “it is important to me that people do not think I am prejudiced” and “when I have negative thoughts about disadvantaged ethnic minorities, I have no concerns expressing them.”
Data Analysis
All analyses will be conducted with STATA 14 (TIDES Protocol, 2017). Descriptive statistics will help display the prevalence of experienced discrimination in relation to the socio-demographic and socioeconomic characteristics of the data sample. A multivariate analysis of variance (MANOVA) will be implemented to analyse the relationship between experienced discrimination in the workplace and both internalised discrimination and explicit bias as outcomes. The dichotomous dependent variable will be coded as 0 and 1 to increase measure validity. A MANOVA is appropriate as there are multiple dependent variables, as well as a plethora of confounding variables to control.
Methodological Considerations
Because of the cross-sectional study design, no causal inference can be made between experienced discrimination and internalised stigma and explicit bias within HCPs. Although the study can find potential associations, it is important to note the risk of prevalence bias when observing the findings in order to confirm what variable caused the association. Although the use of existing survey data is convenient and low cost, another limitation is the reliability of survey data. Self-reported data can result in bias and higher dropout rates, especially when addressing sensitive topics, which can lead to reverse causality. Survey data is also restrictive when attempting to gather additional details outside the limits of the questionnaire.
A MANOVA to conduct the analyses allows for the measurement of multiple independent and dependent variables simultaneously. This multivariate tool allows for higher control over the multiple confounding or covariate variables within this study including age, ethnicity, gender, migrant status and sexual orientation. However, the sensitivity to potential outliers and the affect of a larger sample size on the assumed homogeneity of the sample must be monitored. These concerns will all be addressed as the study unfolds.
Ethical Consideration
This study has limited ethical considerations. Ethical approval for the TIDES study was received from the King’s College London Research Ethics Committee for Psychiatry, Nursing and Midwifery (reference HR-17/18-4629). The statistical analyses will use pre-collected data and the study calls for no substantial amendments. Data collection follows the General Data Protection Regulations (GDPR), all precautions are taken to protect the anonymity of the participants and no further patient contact is required. All data provided for this study will be coded and kept anonymous throughout the process.
