English-Language Competency Using Written Tests of Health Literacy: RESULTS
A total of 354 patients with Hispanic surnames were approached and asked if they spoke English, Spanish or both, in a convenience sample. Two-hundred-forty-nine patients were Spanish-only speakers, and 105 patients were self-declared bilingual. To ensure that the groups were not substantially different, a demographic comparison of the English speakers and non-English speakers was performed. It demonstrated that there was no significant difference found in the gender or age groups (ps>0.05) but a significant difference for patients with family members in the Spanish-only group (pO.OOl) (Table 1).
One-hundred-five patients (29.7% of 354), all self-proclaimed English speakers, were enrolled in the study. Ninety-eight patients completed the REALM test, and 93 patients completed the STOFHLA test. There were various reasons for not completing both tests, including lack of understanding, left against medical advice, problems with understanding the test, illiteracy and refusal. The enrolled patients’ age ranged from 18-89 years, with the 18-30-age range the most common (48.6%). There were 37.1% males (39/105) and 62.9% females (66/105). The level of education was grade school in 49.0% (50/102) followed by high school in 38.2% (39/102); the rest had completed higher degrees. The only significant difference within the dropout and nonparticipant group was in the variable of whether the patient or their family member was the subject (p=0.008). Within that group who refused, 90% were from patients and 9.2% of refusals from family members. The correlation between REALM and STOFHLA tests was t=7.497, df=3, significance=0.001.
Sixteen of 93 (17.2%) were found to be inadequate on the STOFHLA test; 10.8% (10/93) were found marginal, and 72.0% (67/93) were found adequate. Of those 26 that did not pass, 7.7% (2/26) were deemed by the nurses as not competent, and the rest 92.3% (24 of 26) were deemed as competent. Of those 26 that did not pass the test, the nurses were satisfied with their overall communication at 88.5% (23/26). The physicians determined that 95.5% (64/67) who passed the STOFHLA test were assessed as competent, and 4.5% (3/67) that passed the test were not competent. Of those that did not pass the exam, 84.6% (22 of 26) were judged by the physicians to be competent, with only 7.7% (2/26) not passing both the test and the MD competence assessment.
Table 1. Results
| Spanish-Speaking Only |
English- and Spanish-Speaking |
Significance |
|
| Gender |
P=0.018, dt=l, p=0.893 |
||
| Male | 86/249 (34.5%) |
39/105 (37.1%)’ |
|
| Female | 163/249 (65.5%) |
66/105 (62.9%) |
|
| Age |
P=0.29Z df=4, p=0.893 |
||
| 18-30 | 123/249 (49.4%) |
51/105 (48.6%) |
|
| 31-50 | 80/249 (32.1%) |
41/105 (39.0%) |
|
| 51-70 | 34/249 (13.7%) |
10/105 (9.5) |
|
| 2:71 | 12/249 (4.8%) |
1/105 (1.0%) |
|
| Missing |
2/105 (1.9%) |
||
| Patient | 226/249 (90.8%) |
78/105 (74.3%) |
P= 15.914, df=l, p=0.001 |
| Family member | 23/249 (9.2%) |
26/105 (24.8%) |
|
| Missing |
1/105(1.0%) |
Five of 98 (5.1%) patients tested at third-grade level or below on the REALM test, and 28/98 (28.6%) patients were at the fourth-to-sixth grade level. Of those 33 that did not pass, 12% (4/33) patients were deemed by the nurses as not competent. The physicians determined that 69% (63/98) who passed the test were competent and 2% (2/98) that passed were not competent. Of those that did not pass the exam, 34% (33/98) were judged by the physicians to be competent, and only 2% (2/98) that did not pass the REALM test also failed MD competence assessment. Also, the physicians rated the competency level higher than the test in 26/65 (40%) patients who scored above sixth-grade level. There was no correlation between passing either STOFHLA or REALM and gender, age, education level, parental education or primary language spoken at home (p^0.05).
Table 2. Results
| ANOVA | |||||
| Sum of Squares df Mean Square | F |
Significance |
|||
| Nurse What language does the patient speak | 2.658 | 3 | 0.886 |
1.667 |
0.180 |
| RN Determination of patient competency | 2.810E-02 | со | 9.368E-03 |
0.162 |
0.921 |
| RN What language did you speak with patient | 3.626 | со | 1.209 |
1.834 |
0.148 |
| RN How well do you feel patient understood history | 7.366 | со | 2.455 |
8.375 |
0.000 |
| RN How well do you feel patient understood tests | 3.208 | со | 1.069 |
2.968 |
0.037 |
| RN How well do you feel patient understood diagnosis | 3.573 | со | 1.191 |
2.968 |
0.037 |
| RN How well do you feel patient understood treatment | 3.330 | CO | 1.110 |
2.800 |
0.045 |
| RN Were you satisfied with the overall communication | 5.795 | со | 1.932 |
13.121 |
0.000 |
| MD Determination of patient competency | 0.234 | со | 7.803E-02 |
1.479 |
0.226 |
| MD determinate of patient language | 3.702 | со | 1.234 |
2.029 |
0.116 |
| MD What language did you speak to patient | 11.196 | 3 | 3.732 |
4.058 |
0.010 |
| MD How well did the patient understand history | 7.167 | 3 | 2.389 |
4.646 |
0.005 |
| MD How well did you feel patient understood test | 4.976 | 3 | 1.659 |
3.094 |
0.031 |
| MD How well did you feel patient understood diagnosis | 7.384 | 3 | 2.461 |
3.939 |
0.011 |
| MD How well did you feel patient understood treatment 3.366 | со | 1.122 |
2.189 |
0.096 |
|
| MD were you satisfied with the overall communication | 4.896 | 3 | 1.632 |
9.656 |
0.000 |
Both the nurses and the doctors rated patients similarly within each test. Using the ANOVA test, a significant difference was found between the nurses’ rating of patients’ ability and patients’ actual competency using the STOFHLA test (F=4.344, df=2, p=0.016) and MD rating (p=0.001) (Table 2). This finding was also true using the REALM test, with the nurses showing a significant difference between patients testing competent and nurse assessment (p=.001) and for the MD rating (p=0.001). We did not have a significant number of patients on the lower levels, using both health and educational level tests; most were at the two highest categories within each test. The doctors’ and nurses’ assessments of the patients were concordant with respect to the majority of respondents who took part and had an adequate health and sixth-grade-plus level of understanding. A low level of physician and nurse assessment was not seen with either test; instead, there was a middle-to-high level of physician and nurse concordance that matched the assessment outcomes on both tests.
Category: Health
Tags: English-language competency, interpreters