Table C.
Results of individual reports.
| Report | Interventions / Outcomes | Results |
|---|---|---|
| Anttila et al. (2017) | Mode of professional interpreter, relational and ad hoc interpreter / Satisfaction and Communication. | Satisfaction was highest for prof. video interpreter > interpreter trained physician > IPPI > prof. phone interpreter (P = 0.005). Family member and ad hoc not mentioned. Communication was highest with prof. video interpreter > IPPI > interpreter trained physician > family member > prof. phone interpreter simultaneous > ad hoc > prof. phone interpreter later (P = 0.01). |
| Kuo et al. (1999) | Different types of interpreters used / Satisfaction. | Satisfaction with: Professional hospital interpreter was 92.4% (P = 0.17) Relational interpreter was 85.1% (P < 0.01) Telephone interpreter was 53.3% (P < 0.01) Ad hoc (not physician) was 40% (P = 0.05) |
| Lee et al. (2002) | Type of interpreter compared to language concordant patients / Satisfaction. | No significant difference in satisfaction between language concordant and telephone interpreted consultations. Compared to language concordant patients, use of relational and ad hoc interpreters resulted in lower satisfaction: 54% and 49% vs 77%; P < 0.01 and P = 0.007 respectively. |
| Flores et al. (2012) | Professional, ad hoc or no interpreter / Communication. | Proportion of errors of potential clinical consequence (i.e., communication) was lowest for professional interpreter vs ad hoc and no interpreter. 12% vs 22% vs 20%, respectively (P < 0.01). For mean errors per encounter, there was no significant difference. 32.7 (SD 4.9) vs 33.7 (SD 4.7) vs 32.3 (23.9), respectively. |
| Bischoff et al. (2003) | Professional, ad hoc or no interpreter / Communication. | Percentage of patients reporting physical symptoms: With professional interpreter: 25%, ad hoc interpreter: 26%, and no interpreter: 18% (P = 0.079). Percentage reporting psychological symptoms: With professional interpreter: 32%, ad hoc interpreter: 16%, no interpreter: 18% (P = 0.029). |
| Fagan et al. (2003) | Professional, relational or no interpreter / Clinical outcome. | Compared to patients with no interpreter: Professional telephone interpreter resulted in longer provider times (36.3 min vs 28.0 min (P < 0.001)). As did relational interpreter (34.4 min vs 28.0 min (P < 0.001)). In person professional interpreter did not result in significantly different provider times (26.8 min vs 28.0 min (P = 0.51). In multivariate analysis with no interpreter as reference, professional telephone and relational interpreter resulted in longer mean provider times of 8.3 min [95%CI:3.94;12.7] and 4.58 min [95%CI:1.84;7.33], respectively. |
| Baghci et al. (2011) | Professional or any interpreter / Satisfaction and communication. | Satisfaction in treatment group (i.e., professional interpreter) 96% were "very satisfied" vs. 24% in control group (i.e., any interpreter) (OR = 72 [CI:31;167], p = 0.01). Communication in treatment group 93% found understanding "very easy" vs. 18% in control group (OR = 61 [CI:23;166], p = 0.01). |
| Hampers et al. (2002) | Professional or any interpreter compared to controls / Utilization. | When compared to controls, any interpreter had adjusted results on IV use, admissions and testing (i.e., utilization) of: OR 2.2 CI95(1.2,4.3), OR 2.6 CI95(1.4,4.5), OR 1.5 CI(1.04,2.2), respectively. Professional interpreter compared to controls, on the same parameters: OR 1.2 CI95(0.7,2.1), OR 1.7 CI95(1.1,2.8), OR 0.73 CI95(0.56,0.97), respectively. |
| Jacobs et al. (2007) | Professional or any interpreter / Utilization and satisfaction. | No significant differences in outcome for the groups receiving in-person professional interpreter service, compared to the group receiving any interpreter services (i.e., telephone professional, ad hoc, relational, no interpreter). |
| Flores et al. (2003) | Professional or ad hoc interpreter / Communication. | Number of errors with potential clinical consequences were relatively higher for ad hoc than professional interpreters: 77% vs 53% respectively (P < 0.001), i.e. communication higher for professional interpreters. |
| Gany et al. (2007)A | Professional (three modes) or ad hoc interpreter / Communication. | RSMI produced fewer errors than the other modes of interpretation. Mean linguistic errors per utterance 1.139 (SD=1.737) and 0.019 (SD=0.15) medical errors. With the non-RSMI modes of interpretation there was a 12-fold greater rate of medical errors of moderate or greater significance, per utterance (p = 0.002). |
| Gany et al. (2007)B | Professional (RSMI) or ad hoc interpreter with controls / Satisfaction. | RSMI gave significantly higher Satisfaction. Linear regression of satisfaction with physician communication/care: RSMI mean 0.518, SD 0.351 vs usual methods (i.e., ad hoc interpreter) 0.436, SD 0.330, both with P < 0.05. Controls (language concordant) scored significantly higher on all parameters. |
| Garcia et al. (2004) | Professional or ad hoc interpreter compared to controls / Satisfaction and communication. | On a 100-point scale satisfaction was highest for in-person professional interpreter (mean = 79) compared to ad hoc (mean = 72) and telephone professional interpreter (mean = 74), (P < 0.001). Communication was significantly higher for the in-person professional interpreter group (mean = 78) compared to ad hoc (mean = 71) and telephone professional interpreter (mean = 63), (P < 0.001). |
| Nápoles et al. (2015) | Professional or ad hoc interpreter / Communication. | Adjusted odds of inaccurate interpretation were significantly lower for professional in person (OR=−1.25; 95%CI −1.56,−0.95) and video conferencing (OR=−1.05; 95%CI −1.26,−0.84) than for ad hoc interpreter. |
| Butow et al. (2011) | Professional or relational interpreter / Communication. | Equivalence of communication was achieved by professional interpreters 65% of the time and by relational 50% of the time (95%CI:3%−28% for the difference, P = 0.02. |
| Xue et al. (2019) | Professional or relational interpreter / Communication. | Communication presented as concordance on a scale from 0 to 100 divided in intervals. Difference in concordance when comparing professional to relational interpreter was minimal with kappa = 0.69–0.87 and ICCs above 0.74, i.e., gave equal communication results. |
| Hartford et al. (2019) | Professional or no interpreter compared to language concordant patients / Clinical outcome. | Clinical outcome may have been affected negatively for patients with no interpreter, as they had lower chance of ED admittance, but higher risk of ICU admittance within 24 h of first visit, when compared to patients with professional interpreter service or language concordant patients. |
| López et al. (2015) | Professional or no interpreter / Clinical outcome and utilization. | Patients with no interpreter or professional interpreter with a non-physician (i.e., nurse) had significantly shorter stays, OR of 0.80 and 0.77 respectively. There were no significant differences in use of ED and readmission between groups. |
| Luan Erfe et al. (2017) | Professional or no interpreter / Utilization. | Patients with no interpreter were significantly less likely to receive defect-free care (i.e., fully utilized care) compared to with a professional interpreter: 61.5% vs 73.9%, P = 0.04. After accounting for sociodemographics patients with no interpreter were half as likely to receive defect-free care, compared to with a professional interpreter: OR 0.50, 95%CI(0.27–0.90), P = 0.02. |
| Lindholm et al. (2012) | Professional or not interpreter / Clinical outcome. | Length of stay for a patient with professional interpreter at both admission and discharge was 2.57 days while 5.06 days for patients no interpreter (P<0.001). Readmission within 30 days were 24.3% for patients without interpreter present at admission and discharge compared to patients with interpreter present at both 14.9% (P<0.001). |
| Baker et al. (1996)* | Any or no interpreter / Communication. | Communication with any interpreter used: 57% with good-excellent understanding of diagnosis, 43% fair-poor (P<0.001). With interpreter not used: 38% good-excellent understanding of diagnosis, 62% fair-poor (P<0.001). |
| Baker et al. (1998)* | Any or no interpreter / Satisfaction. | On a scale from 0 to 100 patients with an interpreter at visit had an overall satisfaction score of 65 compared to those without interpreter who scored 55 (P<0.001). |
| Bernstein et al. (2002) | Professional or no interpreter with English speaking controls / Utilization. | Patients with no interpreter service provided had less utilization of care than those provided professional interpreter service. Both less than English-speaking patients. (p < 0.05) |
| Moreno et al. (2010) | Any or no interpreter / Satisfaction and communication. | Referenced against patients not needing an interpreter, patients having any interpreter when needed was independently associated with greater satisfaction and communication: 3.65 (SE=1.47) points and 6.04 (SE=1.47) points (P<0.05). Conversely needing an interpreter and not having one showed a decrease in satisfaction and communication: −2.39 (SE=1.15) points and −4.28 (SE=1.42) points (P<0.05). |
| Sarver et al. (2000)* | Any or no interpreter / Utilization. | Referenced to language concordant patients, both patients with interpreter used and not used, were more likely to be discharged without a follow-up appointment. OR 1.92 (1.11;3.33) and 1.79 (1.00;3.23) respectively. P = 0.03. |
| Brooks et al. (2016) | Professional or ad hoc interpreter / Satisfaction, communication and clinical outcome. | Importance of prof. interpreters i.e., ability to relay LEP patients’ medical needs. Barriers to interpretation i.e., time constraints or limited availability of interpreters or use of assumed effective interpreters, i.e., Portuguese interpreters for Spanish speaking patients. Perception of poor care when no interpreter is used, i.e., LEP patients miss crucial information and end up "feeling lost". |
| Greenhalgh et al. (2006) | Professional or relational interpreter / Communication. | Themes identified in relation to distinctions between professional and relational interpreters are their inherent positioning in one of two communicative spheres; 'the system' for the professional and 'the lifeworld' for the relational, sets a basis for the triadic interaction, at the outset. |
| Hilder et al. (2017) | Professional or relational interpreter / Communication. | Patients satisfied with the interpreter used, either professional or family. GPs had differing views; some for professional some for family. Themes were identified as: confidentiality, implicit understanding of the patient's situation, and ability to advocate and assist in the consultation on behalf of the patient. |
| Leanza et al. (2010) | Professional or relational interpreter / Communication. | The study found a higher number of interruptions of the voice of lifeworld (VoL) by physicians with a professional interpreter (64 total) than with a relational interpreter (2 total). VoL was interrupted by relational interpreters more often (21 total) than by professional interpreters (12 total). |
*Based on the same study: conducted at Harbor–UCLA Medical Center, a 500-bed public hospital in Torrance, California, USA.
LEP = limited English proficiency, ED = emergency department, IPPI = in-person professional interpreter, RSMI = remote simultaneous medical interpretation, GP =general practitioner.