PROVIDER REVIEW RECORD · OCTOBER 16, 2026
Lisa Zahakos, RPA-C · Authorized synthetic narration
Technical consistency checks passed. Human pronunciation and clinical release approval remain pending. This record does not certify that every word was pronounced correctly.
All 29 narration sources — 175.3 minutes — received a fresh complete, unprompted local speech-recognition pass. The review also covered all 31 video files: 25 narrated videos matched their intended audio; six visual scene clips are intentionally silent. All 60 files decoded completely, and all 60 live media URLs returned the expected source-file prefix.
Fixed: two obesity-module recordings, “Closing the loop after gestational diabetes” and “The conversation before the next prescription,” had peaks above digital full scale. Both their MP3 files and video soundtracks now have reduced gain. Runtimes, captions, speech content and video imagery are preserved; the corrected files measure below −2 dB true peak. All 54 current audio streams were measured and none exceed digital full scale.
The 131 isolated rechecks included all 29 endings, followed by six additional clarity checks. Sentences that appeared absent in the first recognition pass were recovered in isolated checks; no new script omission was confirmed. Recognition can mishear medicine names and numbers. Seven priority listening checks remain open below.
Still to verify: full human pronunciation approval and actual browser/mobile playback, including seeking. HTTP availability, waveform matching and automatic transcription do not establish these.
The slides summarize the spoken teaching rather than reproduce it word for word. The full captions and transcripts provide the complete written narration. The expanded HFpEF package retains nine baseline, nine formative, nine post and nine reserved follow-up items.
The six episode endings and three full-video endings were re-transcribed, alongside the five flagged passages. No new wording defect was confirmed in that earlier pass. Apparent extra ending words in two automatic transcripts disappeared in isolated, voice-activity-filtered checks and were absent in the earlier independent transcript.
Module 2 of the expanded HFpEF activity now includes a required written dosing note before assessment. The hyperkalemia activity includes a required written clarification about full lessons versus recaps and shared gastrointestinal label cautions. These notes are visible in both learning routes and are not claimed to be spoken in the existing recordings.
The nine identified article and nutrition-wording errors have now been corrected in the recordings, captions and transcripts. Full human listening and pronunciation approval remain pending. The seven priority pronunciation/number-clarity checks below remain open.
Automatic recognition can mishear drug names, acronyms and numbers. These are unresolved listening checks, not established clinical or recording errors. Use the original media below and record the reviewer’s decision in the clinical/media release log.
2:24–3:08 · Listening review pending
Written wording to verify: EMPEROR-Preserved: hazard ratio 0.79; 95% confidence interval 0.69–0.90. DELIVER: hazard ratio 0.82; 95% confidence interval 0.73–0.92.
The recognizer joined the hazard ratio and the following 95% into one numeral. Confirm that their spoken separation is clear.
7:49–8:13 · Listening review pending
Written wording to verify: Spironolactone and eplerenone are steroidal mineralocorticoid receptor antagonists; finerenone is nonsteroidal. Receptor selectivity and pharmacologic properties differ.
The whole sentence is present on targeted re-transcription. Check the pronunciation of the medicine names.
0:49–1:08 · Listening review pending
Written wording to verify: A1C 6.7%; blood pressure 132/80 mm Hg; eGFR 78 mL/min/1.73 m².
The recognizer wrote “130 to 80,” which may represent the homophone “thirty-two/eighty.” The intended systolic value is 132; listening must resolve clarity.
7:20–7:56 · Listening review pending
Written wording to verify: Tirzepatide versus dulaglutide met noninferiority; hazard ratio 0.92, 95.3% confidence interval 0.83–1.01. Superiority was not established.
Confirm both medicine names and the word “noninferiority.” The recognizer returned a phonetic variant; this is a review flag, not a confirmed recording error.
8:34–9:26 · Listening review pending
Written wording to verify: The written passage distinguishes oral from injectable semaglutide, describes the limitations of tirzepatide lactation data, and gives separate Wegovy-tablet and Foundayo wording.
The revised wording is present in the independent transcript. Check semaglutide, tirzepatide, Wegovy, SNAC and Foundayo pronunciation against the manuscript and approved clinical review.
10:39–11:19 · Listening review pending
Written wording to verify: Check the spoken words “secretagogues,” “CYP3A4,” “OATP1B,” and “simvastatin” against the written transcript.
Both the complete-file and isolated unprompted recognition passes returned unclear phonetic variants, particularly for secretagogues. This is a pronunciation review flag, not a confirmed clinical or recording error.
0:35–1:03 · Listening review pending
Written wording to verify: Verify the medicine names Wegovy, tirzepatide and Zepbound in the clinical-conversation recap.
The full-file and isolated recognition passes consistently rendered these names as phonetic variants. Listen to the original before accepting or rerecording the passage.
Full human listening, independent clinical and assessment approval, actual mobile/accessibility playback checks, a timed learner pilot and provider release remain required. The review does not establish accreditation, grant eligibility or upload to myCME.
Host upload guide · Clinical and media review log · Expanded HFpEF activity