HLTH 4900 Week 6 Capstone Synthesis Argument Example

If Translation Helps, Why Do Some Translated Labels Mislead? A Synthesis Built on Verification

Student Name

Walden University

HLTH 4900: Capstone

Instructor Name

Month Day, Year

If Translation Helps, Why Do Some Translated Labels Mislead? A Synthesis Built on Verification

The Tension

Two strands of evidence on translated prescription labels appear to pull apart. One suggests that instructions in a patient's own language, carefully worded, help adults with limited English proficiency understand how to take their medicines. The other documents translated labels, produced by pharmacy software, that mix languages or contain errors. If translation helps, why do some translated labels mislead? This paper does not split the difference between the strands. It looks for the condition under which both can be true and rewrites the capstone claim around it.

The Favorable Strand

The strongest favorable study randomized 202 adults with limited English proficiency, speaking Chinese, Korean, Russian, Spanish or Vietnamese, to review medicine bottles carrying one of two kinds of instructions in their own language: standard translated instructions or a set of standardized, plainly worded instructions designed with health literacy principles (Bailey et al., 2012). Understanding was measured by asking participants to demonstrate how they would take the medicines, not only to state the instructions. Those who received the standardized instructions showed significantly better understanding and dosing. The study's comparison matters: both groups read their own language, so the result shows that the quality of a translated instruction affects understanding, not that translation beats English.

A large population survey supplies the other half of the favorable case. Using California health interview data, researchers found that limited English was associated with roughly three to four times the odds of reporting difficulty understanding prescriptions among Mexican, Central American, Korean and Chinese respondents, though not among Vietnamese respondents (Masland et al., 2011). Understanding there was self-reported, not tested, but the pattern makes clear that language is a barrier translated instructions could address.

The Error Strand

The error strand comes mainly from a survey of pharmacies in the Bronx, New York. Of 286 participating pharmacies, about three-quarters provided labels in Spanish, and most of those used computer programs to produce them. When researchers examined 76 Spanish labels generated by 13 programs, 43% contained incomplete translations mixing English and Spanish, and with misspellings and grammar errors included, half of the labels had errors (Sharif & Tse, 2010). The authors called the quality inconsistent and potentially hazardous.

This strand measured the accuracy of labels, not whether patients understood them. No patient was asked to read or demonstrate a Bronx label. The two strands therefore measured different things: one tested understanding of carefully prepared translations, the other inspected the translations pharmacies actually produce. Treating them as rival answers to one question is the mistake a synthesis has to avoid.

The Condition That Lets Both Hold

Read together, the strands do not conflict. The favorable strand shows that a translated instruction helps when its wording is clear, standardized and accurate, because that is what the tested instructions were: written and checked by researchers before patients saw them. The error strand shows that the translations pharmacies produce in practice are often not accurate, because software generates them and no one checks them. The condition that reconciles the two is verification: translation helps when the translated text has been checked for accuracy and clarity, and the evidence suggests that check is what routine pharmacy practice lacks.

Neither source says this. The randomized study does not discuss software errors, and the Bronx survey does not test understanding. The statement emerges only from setting them side by side and asking why their implications differ.

The condition also explains a detail that would otherwise be puzzling. In the randomized study, even the comparison group read instructions in its own language, yet understood them less well than the group given standardized wording (Bailey et al., 2012). If translation alone were enough, the two groups would have performed alike. That they did not shows that what a translated label says, and how carefully it was prepared, matters as much as the fact of translation. The Bronx findings describe what happens when that preparation is skipped.

The Claim Rewritten

The outline's working claim said that translated labels with plain, standardized wording can help. The synthesis adds one more requirement. Revised claim: Labels in the language a patient reads best, plainly and consistently worded and checked for accuracy, can help adults with limited English proficiency grasp their dosing instructions at community pharmacies. Checked for accuracy turns the error strand from a threat into a requirement. The requirement also points toward practice: a pharmacy that wants its translated labels to help needs a way of checking them, such as standardized instruction sets translated and reviewed in advance rather than generated anew for each prescription.

Why Not Split the Difference

A compromise, such as translated labels sometimes help and sometimes mislead, would be accurate and useless. It would not tell a pharmacy what to do, and it would misread both strands by treating them as measuring the same thing with different results. Nor would it be right to read the error strand as proof that translation fails; the Bronx labels were never tested on patients. The condition is more useful than either a compromise or a verdict because it says when translation helps and which safeguard makes the difference.

What Neither Strand Measured

A pharmacy survey from Milwaukee adds context without changing the synthesis: 47% of the pharmacies that answered said they were never or only sometimes able to produce a non-English label (Bradshaw et al., 2007). Where translated labels are unavailable, the verification question does not yet arise; where they are available, it decides whether they help.

Neither strand asked patients how it felt to receive a translated label, whether they trusted it, whether a label with errors damaged their confidence in the pharmacy or whether they preferred a translated label to an explanation at the counter. Those experiences could affect whether people read the label at all, and a pharmacy that printed accurate labels nobody trusted would see little benefit. The capstone cannot fill that gap from the literature, but it will name it as a question for future work in the limitations section.

References

Bailey, S. C., Sarkar, U., Chen, A. H., Schillinger, D., & Wolf, M. S. (2012). Evaluation of language concordant, patient-centered drug label instructions. Journal of General Internal Medicine, 27(12), 1707-1713. https://doi.org/10.1007/s11606-012-2035-3

Bradshaw, M., Tomany-Korman, S., & Flores, G. (2007). Language barriers to prescriptions for patients with limited English proficiency: A survey of pharmacies. Pediatrics, 120(2), e225-e235. https://doi.org/10.1542/peds.2006-3151

Masland, M. C., Kang, S. H., & Ma, Y. (2011). Association between limited English proficiency and understanding prescription labels among five ethnic groups in California. Ethnicity and Health, 16(2), 125-144. https://doi.org/10.1080/13557858.2010.543950

Sharif, I., & Tse, J. (2010). Accuracy of computer-generated, Spanish-language medicine labels. Pediatrics, 125(5), 960-965. https://doi.org/10.1542/peds.2009-2530