There is a strong correlation between the language proficiency of doctors and the quality of health care patients receive.¹ The National Health Act 61 of 2003 states that health providers ought to communicate health-related matters with their patients in a language that the patients understand and in a manner that considers the patients’ level of literacy.²

Misunderstanding between doctor and patient because of limited language concordance leads to misdiagnosis of diseases and decreases the likelihood of patients to adhere to treatment regimens; this also decreases the patient’s trust in the doctor, eventually leading to an ineffective treatment regimen.^3,4,5 Some patients have reported receiving minimal attention because of their inability to communicate with their doctors fluently, compromising their safety, comfort and satisfaction.^3,4,6,7 A study from Boston, United States (US), found that patients with low English proficiency were more likely to have poorly controlled hypertension than those with adequate English proficiency.⁸ Levin emphasised in his literature review that language difficulty resulted in a decreased understanding of diagnoses, medication and follow-up, as well as decreased adherence to medical advice.⁹ His research found that most patients believed that language and cultural barriers were the most significant impediment to healthcare.⁹

Patients’ ability to comprehend the health language, their adherence to curative care and satisfaction have been shown to improve with language concordance.³ Language concordance reduced language-related misconceptions.¹⁰

A study from California, US, found that the increased effectiveness of the treatment of patients with diabetes resulted from an improved patient-practitioner relationship aided by the physicians’ high proficiency in the patients’ language.¹⁰ Studies in the US regarding Spanish-speaking patients reported enhanced cooperation and an increased likelihood of patients sharing relevant diagnostic information when treated in their own language.^11,12

Poor local language proficiency also has consequences for doctors. It can affect their confidence, competence and work satisfaction. The study from California found that linguistic impediments reduce health workers’ work effectiveness and ability to make holistic care available, as it consumes more time to make a precise medical diagnosis, which leads to frustration.¹⁰ Another US study found misdiagnosis as a result of language discordance.¹³ A study from the United Kingdom reported that in a clinical setting, healthcare professionals felt that the most problems arose from them not understanding idioms and colloquial language used by patients. Such misunderstandings could have adverse effects on patient safety.¹⁴

Doctors in the Free State province, South Africa, lacked adequate language proficiency in Sesotho, the most common local language, according to a 2006 study,¹⁵ despite most respondents being trained in the province. The study recommended better language training during the undergraduate medical programme.¹⁵

Language proficiency is not only a problem after graduation. Clinical training for medical students can be challenging if they lack language proficiency. A study from Qatar¹⁶ showed that healthcare students who were less diverse in language skills faced more frustration and stress in the clinical setting. Abdelrahim et al.¹⁶ also reported that language discordance could lead to lower quality of health information, higher cost implications and increased health risks. Therefore, providing comprehensive language teaching for health students in clinical training is essential. A study¹⁷ from KwaZulu-Natal, South Africa, found that teaching students introductory isiZulu in their first year enhanced their training experience. Even though they could not communicate fluently in isiZulu by their third year, they were still positive and wanted more language training.

Medical schools have the responsibility to adapt their curriculum to prepare students for the language challenges they will face in their clinical years and beyond. However, before making any changes, the proficiency levels of students need to be assessed.

Of the 119 fourth-year medical students at the UFS, 96 completed the self-administered questionnaires (response rate 80.7%). Of the 96, 41 (42.7%) consented to the telephonic interviews. Of these 41 students, 28 (68.3%) answered when contacted to participate in the telephonic interviews. As shown in Table 1, there were more females in the class than males, and females were the majority of the participants in the two phases. The median age of the participants was 23 years (range 21–34 years).

In Table 2, the predominant home language (the primary language spoken at home) among the participants was Afrikaans (33.3%), followed by English (25.0%) and Sesotho (20.8%). English was the predominant first language among the study participants (62.5%), followed by Afrikaans (32.3%) and Sesotho (2.1%). The predominant first additional language among the study participants was Afrikaans, followed by English and Sesotho. The ‘other languages’ are defined as languages not learned as part of the school curriculum or spoken at home as the primary language.

Of the 37 participants who indicated how they learned their first additional language, family was the most common source (32.4%), followed by school (29.7%), ‘picked it up’ (21.6%) and university (18.9%). The remaining participants learned from friends and television.

The largest percentage of the study participants in Phase I rated themselves as beginners or elementary (54.2%) in Sesotho, whereas the largest percentage rated themselves as either advanced or proficient (47.9%) in Afrikaans (Table 3). Participants in Phase II were more likely to rate themselves as proficient in Sesotho and less likely to rate themselves as proficient in Afrikaans than those participating in Phase I.

Table 4 indicates the scores for the individual questions of the interview assessment. The median overall scores on the scale from 4 to 16 were 9 (Sesotho) and 10.5 (Afrikaans) with IQR 4 to 16 for both languages. Eight students (28.6%) were completely proficient (scored 16/16) in Sesotho and eight (28.6%) in Afrikaans, whereas 10 (35.7%) had no knowledge of the language (scored 4/16) in Sesotho and 10 (35.7%) in Afrikaans. One participant scored 4 in both languages, and the most proficient student scored 15 and 16 in the two languages, respectively.

Table 5 summarises the actual language proficiency tested in the interview, grouped according to the students’ self-rated language proficiency. Students with higher self-rated language proficiency had higher actual language proficiency scores. The Spearman rank correlation coefficients between self-rated language proficiency and actual language proficiency score were 0.93 for Sesotho and 0.96 for Afrikaans.

The high response rate in Phase I and the gender distribution, which reflects the study population, increase the representativeness of the participants from Phase I. The participants had diverse home languages, including Sesotho, Afrikaans, English and other official languages. However, the majority received their school education in English. Most of those with an additional language learned from their family, school or ‘picked it up’ in the community. Only a small percentage (18.9%) learned an additional language informally while at university.

Although Phase II of the study had a much lower response rate, the study confirmed that students have insight into their own language capabilities. This insight on the part of the students should enhance attempts to improve students’ language proficiency. In a study of final-year medical students at the University of KwaZulu-Natal, South Africa, students were predominantly positive regarding the need for doctors to be at least bilingual.²⁰ A different study by the same research team found that students in their third year had positive attitudes regarding communication competence 2 years after the introductory training of a year-long module in isiZulu, even though they did not yet feel competent in having conversations in the additional language.¹⁷ Students are aware of their shortcomings and generally are open to additional training.

The language proficiency of the majority of UFS medical students in the most widely spoken languages in the Free State is unacceptably low. Additional training should be implemented in the curricula and not only at an introductory level. Continuous training should be throughout the programme, which could also form part of the assessment in the final exam with a patient from the local communities. Burch has argued that medical curricula emphasise cultural competence but ‘have largely failed to address the challenge of non-English language competence of doctors’. She states that ‘the most basic need of a person seeking healthcare is the need to be heard and understood’.²¹ The use of interpreters would obviously assist with communication between the patient and doctor, but doctors being proficient in the language of their patients can only be advantageous. South Africa, with its 11 official languages (at the time of the study), is a challenging environment as students trained in a specific area, may receive training in languages common in that area (such as Afrikaans and Sesotho in the case of the Free State) but during their internship, community service and later career work in areas where other languages dominate. Adequate language proficiency training can help students become more culturally aware and adaptable to new circumstances.

The authors were able to assess Sesotho and Afrikaans language proficiency in a portion of fourth-year medical students; hence, the aim of the study was accomplished in this regard. Of the sample that underwent the entire course of this study (those whose language proficiency was tested via telephonic interviews), the majority were not proficient in Sesotho and Afrikaans. Only a single participant was assessed as proficient in both Sesotho and Afrikaans. This is a concern as Sesotho and Afrikaans are the most spoken languages in the Free State, making it challenging for medical students to communicate with the patients they encounter during their clinical rotations.

The current language training in Sesotho and Afrikaans in the undergraduate medical training at the UFS does not meet the objective of giving the students the language skills necessary to communicate with the majority of their patients in their home language.