The healthcare inequality experienced by lesbian, gay, bisexual, transgender, queer and other (LGBTQ+) people has been highlighted in the Stonewall (2018) healthcare report, which found that 13% report experiencing unequal treatment from healthcare staff because they are LGBTQ+ and 14% avoided treatment for fear of discrimination because of belonging to this group.
A recent report from LGBTQ+ Health and Wellbeing, a charity in Scotland, had similar findings, highlighting health inequality among LGBTQ+ people. It said: ‘LGB people are twice as likely to report symptoms of poor mental health’ and ‘have 1.5 times higher prevalence of depression and anxiety disorders than heterosexual adults; LGB people aged 55+ showed twice the prevalence of poor mental health’ (Semlyen et al, 2016). It found similar statistics of high rates of anxiety, depression and self-harm in young LGBTQ+ people, with higher rates among transgender people of all ages. This report discusses healthcare providers' attitudes and perceptions towards LGBTQ+ people, noting that ‘22% of LGBTQ+ people felt uncomfortable being open about their sexual orientation or gender identity, rising to 33% for people accessing social care’ (LGBTQ+ Health and Wellbeing, 2018).
Robinson (2019) suggested that transgender people were reluctant to access primary care services such as a GP, citing a government survey of >100 000 LGBTQ+ people, highlighting a lack of understanding of gender identity and referral pathways as barriers to accessing healthcare. Robinson (2019) also discussed another survey with more than >30 000 LGBTQ+ respondents, with 20% reporting feeling healthcare services were unsupportive when they were trying to access services.
An earlier study by Sharek et al (2014) found similar healthcare inequality in Ireland. This study focused on elderly LGBTQ+ people. While it discussed positive examples of healthcare interactions, it found that only 43% of patients felt respected as an LGBTQ+ person by health professionals and 26% chose not to disclose their LGBTQ+ status for fear of a negative response.
Healthcare equality regarding LGBTQ+ people has been acknowledged by the NHS; a committee was formed to develop a strategy focused on improving access to and uptake of healthcare by LGBTQ+ people. Recommendations from the strategy included improving the patient experience by making the services more visibly LGBTQ+ friendly, improving training for staff in supporting LGBTQ+ patients and monitoring patient demographics to help the services meet the previous two recommendations (NHS Equality and Diversity Council, 2015). Studies since the above suggest that progress is limited, and more work is needed to reduce healthcare inequalities.
A House of Commons Women and Equalities Committee report (2019) addressing health and social care and LGBTQ+ communities highlighted ‘unacceptable inequalities in health outcomes’ and stated: ‘Monitoring both sexual orientation and gender identity is far too important to be an aspiration rather than a concrete goal.’ This suggests that data collection is strategically a high priority but it is not known if this information is recorded at an operational level (House of Commons Women and Equalities Committee, 2019).
A study was carried out to evaluate the sexual orientation and gender identity of patients presenting to the ambulance service to determine whether there were any trends in presentation and if data was being collected accurately and to attempt to compare this to nationally collected data.
No literature relates to health inequality relating specifically to patients of the ambulance service. This could suggest that such inequalities were not experienced from the ambulance service, or that patients were not expressing their sexual orientation and gender identity or not being asked.
This is despite both the NHS Equality and Diversity Council (2015) and the House of Commons Women and Equalities Committee (2019) strategies highlighted the need to record the gender identity and sexual orientation of patients to support data gathering and improve health inequalities for LGBTQ+ people.
Methods
Four out of the 10 ambulance services in England were chosen for inclusion in the study to sample different geographic locations with a mix of environments and population densities (such as urban and rural areas). Each service was contacted with a request to disclose information under the Freedom of Information Act. The following information was requested for the years 2017, 2018 and 2019:
The ambulance services were also asked to provide data on paper and electronic records separately to determine if there was any disparity dependent on the recording means used. A list of potential responses to the above questions was also requested.
The Office for National Statistics (ONS) gathers data on the sex, sexual orientation and gender identity of the general public in the UK. These data are now available online and were collected for the same period (though as mentioned, were not available at the time of writing).
Results
All four of the services contacted responded. One provided data with an incorrect date range and in a financial year format; this service's data were excluded from the study. However, in its response, this service said its clinical records system did not have the option to record a patient's sexual orientation. The remaining services were labelled services 1, 2 and 3.
Service 1 provided data for gender identity recorded in the requested time range (Table 1); it stated that it did not collect data on sexual orientation. This service did not provide separate data for paper or electronic clinical records, nor did it specify whether it used either or both of these. Data from this service was tabulated, and a minimal variation in gender identity was observed across the sample. Annually, between 10.30% and 11.63% of patients were recorded as ‘unknown’; it is not known why this option was selected.
| 2017 | 2018 | 2019 | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Service 1 | Service 2 | Service 3 | Service 1 | Service 2 | Service 3 | Service 1 | Service 2 | Service 3 | ||||||||||
| Female | 276 176 | 47.3% | 167 430 | 51.4% | 240 566 | 53.0% | 279 612 | 47.2% | 168 680 | 50.0% | 272 465 | 53.0% | 241 575 | 47.0% | 178 455 | 50.4% | 291 581 | 53.2% |
| Male | 247 836 | 42.4% | 150 637 | 46.3% | 211 283 | 46.5% | 250 713 | 42.3% | 158 931 | 47.1% | 239 067 | 46.5% | 212 363 | 41.3% | 166 199 | 47.0% | 252 651 | 46.1% |
| Non-binary | 10 | 0.0% | 10 | 0.0% | 11 | 0.0% | ||||||||||||
| Trans male | 46 | 0.0% | 61 | 0.0% | 53 | 0.0% | ||||||||||||
| Trans female | 30 | 0.0% | 41 | 0.0% | 46 | 0.0% | ||||||||||||
| Unknown | 60 139 | 10.3% | 7349 | 2.3% | 2325 | 0.5% | 62616 | 10.6% | 9441 | 2.8% | 2143 | 0.4% | 59 740 | 11.6% | 9062 | 2.6% | 3889 | 0.7% |
| Total | 584 151 | 325 502 | 454 174 | 592 941 | 337 164 | 513 675 | 513 678 | 353 826 | 548 121 | |||||||||
Data provided by service 2 (Table 1) included both electronic and paper records; the number of electronic records significantly exceeded that of paper records; for this reason, comparison between electronic and paper records was not completed. However, it was noted that options for responses for paper and electronic records differed; electronic records had the option to record a patient as non-binary whereas paper records did not.
This service had the option for the patient to record their gender identity as non-binary, trans male or trans female. However, only a fraction (less than 0.05%) of patients presented within these categories. Minimal variation was observed between responses across the sample period.
Data from service 2 also found that some patients were recorded as ‘unknown/not specified’, with this group ranging between 2.30% and 2.80% annually between 2017 and 2019. Service 2 also reported that its clinical records did not have the option to record patients' sexual orientation.
Service 3's (Table 1) data also separated records between paper and two electronic systems in use; because the majority of records were on one electronic system, the data were analysed as a collective group. It was also found that different options were available on paper and the two types of electronic clinical record. Minimal variation in gender identity was observed across the sample period. Service 3 did have the option to record a patient's sexual orientation on cliwnical records. It was again found that potential responses varied between different systems as seen in Table 2.
| Paper | Electronic A | Electronic B |
|---|---|---|
| Bisexual | Bisexual | Bisexual |
| Gay | Gay | Homosexual |
| Lesbian | Lesbian | |
| Heterosexual | Heterosexual | Heterosexual |
| Not recorded | Not recorded | Not recorded |
| Refused/declined | Refused/declined | Refused/declined |
| Not stated | ||
| Patient unable [to provide] |
Gay and lesbian were grouped as homosexual during data review from service three (Table 3). There was little variation in responses across the three sample years other than a decrease in patients recorded as ‘not stated’ and an increase in those marked as ‘patient unable’ [to provide] or ‘refused/declined’. Other responses varied marginally across the sample period but were massively outweighed by responses such as ‘not recorded’, ‘not stated’ or ‘refused/declined’, which represented between 83.13% and 94.97% of data across the sample period.
| Sexual orientatin | 2017 | 2018 | 2019 | |||
|---|---|---|---|---|---|---|
| Bisexual | 281 | 0.06% | 151 | 0.03% | 169 | 0.03% |
| Homosexual | 418 | 0.09% | 542 | 0.11% | 575 | 0.10% |
| Heterosexual | 22135 | 4.87% | 25 400 | 4.94% | 37 720 | 6.88% |
| Not recorded | 251 614 | 55.40% | 261 607 | 50.93% | 276 238 | 50.40% |
| Not stated | 103 916 | 22.88% | 148 228 | 28.86% | 73 039 | 13.33% |
| Patient unable [to provide] | 0 | 0.00% | 0 | 0.00% | 54 052 | 9.86% |
| Refused/declined | 75 810 | 16.69% | 77 747 | 15.14% | 106 328 | 19.40% |
| Total | 454 174 | 513 675 | 548 121 | |||
The ONS collects and produces data relating to the sexual orientation of the public based on data gathered from annual population surveys (Table 4); data for 2019 were not available at the time of analysis (ONS, 2020).
| Sexual orientation | 2017 | 2018 |
|---|---|---|
| Heterosexual or straight | 95.0% | 94.6% |
| Gay or lesbian | 1.3% | 1.4% |
| Bisexual | 0.8% | 0.9% |
| Other | 0.6% | 0.6% |
| Do not know or refuse | 2.3% | 2.5% |
Source: Office for National Statistics (2019)
The ONS (2020) also gathers information about the gender identity of the UK population; data were not yet available for 2018 and 2019. Regarding sex, it was estimated in 2017 that the proportions of women and men in the UK populations were 50.66% and 49.34% respectively.
Discussion
Analysis of the data provided highlighted several themes relating to how personal characteristics of patients were being recorded. These concerned:
Limited and inconsistent recording of gender identity
Services 1, 2 and 3 all reported instances of a patient's gender identity as ‘unknown/not specified’ or similar terms, which reflect gaps in data capture. The percentage of patients with an ‘unknown’ record was consistent annually within individual services but varied between them.
The proportion of patients recorded as ‘unknown’ was considerably higher for service 1. This may be because services 2 and 3 used an electronic clinical record and the higher unknown rate in service 1 could be attributed to transcription errors or fields being left blank; it may also be possible that electronic records are automatically populated without user input or are required fields, which must be filled in order for the clinical record to be completed. More discussion with the clinicians from service 1 may reveal alternative reasons for the high rate of data being omitted.
Despite absent data, it was found across all three services that the proportion of male and female patients was consistent with ONS (2020) data.
Service 2 had the facility to record the demographics of non-binary and transgender patients, although these groups accounted for only a small proportion of the sample; this may be because of data being recorded incorrectly or patients identifying as male or female rather than trans male or trans female.
Limited and inconsistent recording of sexual orientation
Out of the four services contacted, only one could record patients' sexual orientation despite individual NHS and parliamentary strategies highlighting the need to record these demographics to support monitoring LGBTQ+ patients' use and experience of healthcare. As data sampling ended in 2019, it is possible that has since been implemented as part of individual services' own development. Nonetheless, service 3 has had the facility to record a patient's sexual orientation since at least 2017.
Only service 3 had the facility to record the sexuality of a patient. The review of data from this service showed that the three systems in use had varying options for recording a patient's sexual orientation. One electronic system used by service 3 had options with similar meanings e.g. ‘not recorded’ and ‘not stated’. It is possible that an update in the system being used led to a change in possible responses, which consolidates the list of potential answers, but the scope of this study and the data provided could not be used to confirm this suspicion. Further discussion with the service could clarify the various responses available for a patient's sexual orientation. Different options for sexuality was also observed between the paper and the two computer systems used by this service; this is believed to be the result of a software update mid-way through the sample but has not been confirmed.
From the available data, it was found that the majority of patient sexualities had been recorded as ‘not stated’, ‘not recorded’ or ‘refused/declined’, accounting for 94.97% of patients in the available data; this rate was found to be lower in 2019 (83.13%), but this may be due to 9.86% of patients' sexualities being recorded as ‘patient unable’ in this year. It was found that the rate of gender identity being captured was considerably higher.
Service 3's data showed that the recorded sexualities of patients did not reflect ONS (2019) data. The reason for this could not be explained by this study and could be further explored by a subsequent study.
It is possible that patients are not reporting their sexuality. If this is the case, it should be explored with the patients to determine the underlying reason. It may be that this information is seen as irrelevant by patients seeking healthcare or they may be concerned that their healthcare would be negatively impacted by revealing it.
This may also be because patients are too unwell to respond or asking such a question may have been detrimental to their care; it is also possible that the clinician is not requesting or recording this information or patients are refusing to provide it.
Clinicians could be not requesting or recording this information for reasons such as not believing it is relevant or not feeling comfortable asking for it.
The difference between gender identity and sexual orientation data being captured could not be explained. Several possible causes are discussed but the scope of this study could not support a conclusion being drawn, and further research is needed to determine why information on sexual orientation is not being captured and how the capture rate could be improved.
Completeness of data
The data recorded for patient gender identity appear to reflect ONS (2020) data, but the same was not found for the data relating to patient sexuality. Therefore, it is possible the data could not be considered accurate because of the amount of absent data. Further research to establish ways to improve the capture and quality of data that may help produce a more representative sample has already been recommended.
Data recording between services
Variations in the rate of data being captured were found relating to gender identity and sexual orientation between all three services. The three services had different options for possible responses to gender identity; service 3 had different possible responses regard sexuality in each of its systems.
This lack of consistency between services will have a detrimental impact on data quality and negatively impact the value of reporting from the resulting inability to match data between organisations.
Agreeing on a standardised set of responses for all ambulance services to use in combination with a strategy of improving data capture would support the creation of a generalisable data set to support government and NHS objectives (NHS Equality and Diversity Council, 2015; House of Commons Women and Equalities Committee, 2019) to improve care for LGBTQ+ people.
Limitations
The study gathered data over 3 years to highlight any variation, and was able to compare this information both within and between services (although this turned out not to be the ultimate aim of the study). Corresponding data from the ONS were not available for 2018 and 2019 at the time of the study, so no further comparison could be drawn, which has potentially limited the study's impact.
Each service had a different format for recording data, with different terms for missing data such as ‘refused’ or ‘not recorded’; these options may not have been directly transferable and may have impacted the quality of data used in the study.
Similarly, data relating to sexual orientation were not available from the ONS. Several charities were approached to determine whether a data set was available to allow comparison, but these data would not have been a representative sample of the population of the UK. These data became available only after the 2021 census, which took place since this study was carried out. Also, data relating to sexuality were collected only by one of the four services contacted; this prevented the study from reviewing any variation in recorded data.
This study contacted four ambulance services in England and analysed data from three; this produced a data sample for analysis but it may not be truly reflective of the national picture as the UK has 13 ambulance services.
It is also possible that processes have changed since the data were collected and these demographic details may now be collected more accurately.
Conclusion
The data and subsequent analysis in this study have gained an insight into the way gender identity and sexuality are recorded by ambulance services in England as well as some flaws and gaps in recording.
The format for recording gender identity differs between services; their rates of missing data vary considerably. It is possible that the format or method of recording may be a factor in this missing data. Only one service has made a provision for recording the gender identity of non-binary and transgender patients.
Only one ambulance service had mechanisms to record the sexual orientation of a patient and the format varied between the three systems in place, It was therefore not possible to compare statistics between services. In addition, the amount of data on sexual orientation was greatly exceeded by the amount of data that was absent. In addition, there was disproportionately more data missing on sexual orientation than on gender identity.
The recording of gender identity varied between services and it appears that sexual orientation is less likely to be recorded and may not be recorded accurately.
One theme the study can confirm is that systems in place to record data are poor; it is recommended that the UK ambulance services agree on a format for recording a patient's gender identity and use this across all recording systems to allow a more consistent data set to be produced. The same would be recommended regarding sexual orientation for all services. Those that are not recording these data should be offered support to do so.
This study should be extended to include the remaining ambulance services in the UK to determine if similar findings of recorded patient demographics are observed. A qualitative study involving patients and or clinicians may also be beneficial to help identify any factors that impact the recording of data and subsequent data quality, including when would be the most appropriate time to collect these demographics.