{
  "docId": "019dd923-622c-750b-8b9a-21918ac10c17",
  "docSlug": "f3db2294dcf9",
  "documentTitle": "Harmony Biosciences (HRMY)",
  "authorId": "58_Scorpion_Capital",
  "authorName": "Scorpion Capital",
  "documentKindSlug": "research-note",
  "documentKindLabel": "Research note",
  "sourceTypeSlug": "short_seller",
  "sourceTypeLabel": "Short seller",
  "presentationDate": "2023-03-28 00:00:00",
  "orientation": "landscape",
  "aspectRatio": 1.3333334,
  "pageNumber": 224,
  "pageCount": 366,
  "prevPage": 223,
  "nextPage": 225,
  "slideType": "expose_contradiction",
  "function": "expose_contradiction",
  "density": "dense",
  "nDataPoints": 2,
  "notes": "The slide uses a 'red flag' argument to suggest data manipulation or p-hacking via sample size adjustment.",
  "elementsJson": [
    "paragraph",
    "callout_box",
    "screenshot"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-622c-750b-8b9a-21918ac10c17/224",
  "deckHref": "/decks/019dd923-622c-750b-8b9a-21918ac10c17",
  "deckJsonHref": "/decks/019dd923-622c-750b-8b9a-21918ac10c17.json",
  "deckAnchorHref": "/decks/019dd923-622c-750b-8b9a-21918ac10c17#slide-224",
  "components": [
    {
      "bbox": {
        "h": 0.03,
        "w": 0.86,
        "x": 0.07,
        "y": 0.76
      },
      "kind": "callout",
      "text": "However, the history of changes on ClinicalTrials.gov indicates the trial size doubled midway - excerpt",
      "attrs": null,
      "subkind": "primary",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "8972602c-8daa-468d-b818-12e4349c424a",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.03,
        "w": 0.86,
        "x": 0.07,
        "y": 0.5
      },
      "kind": "callout",
      "text": "Harmony 1 trial paper makes no mention of trial expansion and makes the n of 100 appear pre-specified",
      "attrs": null,
      "subkind": "primary",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "8b7713be-4629-4514-812d-164f58daaf50",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": null,
      "kind": "callout",
      "text": "Mid-trial study adjustments which expand the sample size are a red flag and predictive of a flop, as they suggest that the study is failing to show statistical significance and using a larger n to fish for a smaller effect.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd953-64d1-73a9-abb6-97d4490883c4",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.12,
        "w": 0.55,
        "x": 0.07,
        "y": 0.8
      },
      "kind": "image",
      "text": "ClinicalTrials.gov archive\nHistory of Changes for Study: NCT01067222\nEfficacy and Safety Study of BF2.649 in the Treatment of Excessive Daytime Sleepiness in Narcolepsy (Harmony1)\nEnrollment: 60 [Anticipated] 110 [Actual]",
      "attrs": null,
      "subkind": "screenshot",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "a5ba4fb0-89ca-4d0c-bbb1-3b81c2c46d50",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.42,
        "w": 0.86,
        "x": 0.07,
        "y": 0.07
      },
      "kind": "paragraph",
      "text": "As clinically immaterial as the small ESS reductions in HARMONY 1 are, it appears Bioprojet had to double to trial size mid-way in order to conjure them. The paper makes no mention of a trial expansion, and is written to represent that the final powering of 110 was pre-specified. However, the history of changes for the trial’s record on ClinicalTrials.gov indicates it originally planned to enroll only 60 patients. Mid-trial study adjustments which expand the sample size are a red flag and predictive of a flop, as they suggest that the study is failing to show statistical significance and using a larger n to fish for a smaller effect. Common sense indicates that if a study is pointing to efficacy, one has no incentive to expand the size and rock the boat. Such protocol changes are particularly damning because they mean the sponsor was getting results and modifying the trial with knowledge of day-to-day data.",
      "attrs": null,
      "subkind": "paragraph",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "3fe19570-1879-4558-be6b-278f1a456ac5",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.18,
        "w": 0.38,
        "x": 0.07,
        "y": 0.54
      },
      "kind": "paragraph",
      "text": "Statistical analysis\nWe calculated the sample size on the basis of data from previous trials,20,21 based on the minimum clinically relevant difference on a final ESS of 3 points, ESS SD of 5, and a coefficient of correlation r (baseline ESS, final ESS) of 0·65. The first test (difference in change of ≥3 points detectable with a power ≥95%) and the second test (non-inferiority margin of 2 points and 80% as minimal power) needed a sample size of 30 patients per group.",
      "attrs": null,
      "subkind": "paragraph",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "eeab9028-8191-4bd3-b908-a8c8299a1a20",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.03,
        "w": 0.68,
        "x": 0.07,
        "y": 0.95
      },
      "kind": "source-note",
      "text": "Source: https://pubmed.ncbi.nlm.nih.gov/24107292/; https://clinicaltrials.gov/ct2/history/NCT01067222?A=1&B=4&C=merged#StudyPageTop",
      "attrs": null,
      "subkind": null,
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "7d26e733-1397-4c1f-9c72-682208afac9d",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [],
  "frameworks": [
    {
      "name": "fraud-exposure",
      "slug": null,
      "matchId": "d2690eee-0999-4348-a899-164f227c3e99",
      "evidence": "The slide explicitly frames the mid-trial sample size increase as a 'red flag' and 'damning' evidence of data manipulation.",
      "confidence": 0.9
    }
  ],
  "arcBeats": [],
  "loops": [],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}