Deterioration Index Version 2

{0} FDA U.S. FOOD & DRUG ADMINISTRATION September 18, 2026 Epic Systems Corporation Melissa Knox Official Correspondent 1979 Milky Way Verona, Wisconsin 53593 Re: K260008 Trade/Device Name: Deterioration Index Version 2 Regulation Number: 21 CFR 870.2210 Regulation Name: Adjunctive Predictive Cardiovascular Indicator Regulatory Class: Class II Product Code: QNL Dated: August 20, 2026 Received: August 21, 2026 Dear Melissa Knox: We have reviewed your section 510(k) premarket notification of intent to market the device referenced above and have determined the device is substantially equivalent (for the indications for use stated in the enclosure) to legally marketed predicate devices marketed in interstate commerce prior to May 28, 1976, the enactment date of the Medical Device Amendments, or to devices that have been reclassified in accordance with the provisions of the Federal Food, Drug, and Cosmetic Act (the Act) that do not require approval of a premarket approval application (PMA). You may, therefore, market the device, subject to the general controls provisions of the Act. Although this letter refers to your product as a device, please be aware that some cleared products may instead be combination products. The 510(k) Premarket Notification Database available at https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm identifies combination product submissions. The general controls provisions of the Act include requirements for annual registration, listing of devices, good manufacturing practice, labeling, and prohibitions against misbranding and adulteration. Please note: CDRH does not evaluate information related to contract liability warranties. We remind you, however, that device labeling must be truthful and not misleading. If your device is classified (see above) into either class II (Special Controls) or class III (PMA), it may be subject to additional controls. Existing major regulations affecting your device can be found in the Code of Federal Regulations, Title 21, Parts 800 to 898. In addition, FDA may publish further announcements concerning your device in the Federal Register. Additional information about changes that may require a new premarket notification are provided in the FDA guidance documents entitled "Deciding When to Submit a 510(k) for a Change to an Existing Device" U.S. Food & Drug Administration 10903 New Hampshire Avenue Silver Spring, MD 20993 www.fda.gov {1} K260008 - Melissa Knox Page 2 (https://www.fda.gov/media/99812/download) and "Deciding When to Submit a 510(k) for a Software Change to an Existing Device" (https://www.fda.gov/media/99785/download). Your device is also subject to, among other requirements, the Quality Management System Regulation (QMSR) (21 CFR Part 820), which includes, but is not limited to, ISO 13485 clause 7.3 (Design controls), ISO 13485 clause 8.3 (Nonconforming product), ISO 13485 clause 8.5.2 (Corrective action), and ISO 13485 clause 8.5.3 (Preventative action). Please note that regardless of whether a change requires premarket review, the QMSR requires device manufacturers to review and approve changes to device design and production (ISO 13485 clause 7.3 and ISO 13485 clause 7.5) and document changes and approvals in the Medical Device File (ISO 13485 clause 4.2.3). Please be advised that FDA's issuance of a substantial equivalence determination does not mean that FDA has made a determination that your device complies with other requirements of the Act or any Federal statutes and regulations administered by other Federal agencies. You must comply with all the Act's requirements, including, but not limited to: registration and listing (21 CFR Part 807); labeling (21 CFR Part 801); medical device reporting (reporting of medical device-related adverse events) (21 CFR Part 803) for devices or postmarketing safety reporting (21 CFR Part 4, Subpart B) for combination products (see https://www.fda.gov/combination-products/guidance-regulatory-information/postmarketing-safety-reporting-combination-products); good manufacturing practice requirements as set forth in the Quality Management System Regulation (QMSR) (21 CFR Part 820) for devices or current good manufacturing practices (21 CFR Part 4, Subpart A) for combination products; and, if applicable, the electronic product radiation control provisions (Sections 531-542 of the Act); 21 CFR Parts 1000-1050. All medical devices, including Class I and unclassified devices and combination product device constituent parts are required to be in compliance with the final Unique Device Identification System rule ("UDI Rule"). The UDI Rule requires, among other things, that a device bear a unique device identifier (UDI) on its label and package (21 CFR 801.20(a)) unless an exception or alternative applies (21 CFR 801.20(b)) and that the dates on the device label be formatted in accordance with 21 CFR 801.18. The UDI Rule (21 CFR 830.300(a) and 830.320(b)) also requires that certain information be submitted to the Global Unique Device Identification Database (GUDID) (21 CFR Part 830 Subpart E). For additional information on these requirements, please see the UDI System webpage at https://www.fda.gov/medical-devices/device-advice-comprehensive-regulatory-assistance/unique-device-identification-system-udi-system. Also, please note the regulation entitled, "Misbranding by reference to premarket notification" (21 CFR 807.97). For questions regarding the reporting of adverse events under the MDR regulation (21 CFR Part 803), please go to https://www.fda.gov/medical-devices/medical-device-safety/medical-device-reporting-mdr-how-report-medical-device-problems. {2} K260008 - Melissa Knox Page 3 For comprehensive regulatory information about medical devices and radiation-emitting products, including information about labeling regulations, please see Device Advice (https://www.fda.gov/medical-devices/device-advice-comprehensive-regulatory-assistance) and CDRH Learn (https://www.fda.gov/training-and-continuing-education/cdrh-learn). Additionally, you may contact the Division of Industry and Consumer Education (DICE) to ask a question about a specific regulatory topic. See the DICE website (https://www.fda.gov/medical-devices/device-advice-comprehensive-regulatory-assistance/contact-us-division-industry-and-consumer-education-dice) for more information or contact DICE by email (DICE@fda.hhs.gov) or phone (1-800-638-2041 or 301-796-7100). Sincerely, STEPHEN C. BROWNING -S CDR Stephen Browning Assistant Director Division of Cardiac Electrophysiology, Diagnostics, and Monitoring Devices Office of Cardiovascular Devices Office of Product Evaluation and Quality Center for Devices and Radiological Health Enclosure {3} | Indications for Use | | | | --- | --- | --- | | Please type in the marketing application/submission number, if it is known. This textbox will be left blank for original applications/submissions. | K260008 | ? | | Please provide the device trade name(s). | | ? | | Deterioration Index Version 2 | | | | Please provide your Indications for Use below. | | ? | | Deterioration Index Version 2 is software that calculates the risk of patient deterioration, defined as the patient experiencing an escalation of care (a transfer to the ICU, a rapid response team call, or a code) or mortality. Deterioration Index Version 2 is for use with adult patients aged 18 or older who are admitted to an inpatient unit, observation unit, or on observation status in the emergency department of a hospital. The software is not for use with patients in a long-term care setting, operating room, post anesthesia care unit (PACU), labor and delivery unit, or for non-observation patients in an emergency department. The software is for use by trained healthcare professionals. Deterioration Index Version 2's risk score is calculated from Epic EHR data, such as vitals that have been documented for the patient. As clinical decision support software, Deterioration Index Version 2's risk score is intended to aid clinicians in identifying which patients are more likely to clinically deteriorate. Deterioration Index Version 2 provides additional information and does not replace the standard of care or clinical judgment. The device is not intended to serve as the primary determinant of health status for a patient, nor as a substitute for a healthcare professional's judgment and decision making. The device is not a primary diagnostic or treatment tool. Information provided by the device is not intended to be relied upon as the sole basis for decisions that have an impact that may cause death or an irreversible deterioration of a person's state of health. | | | | Please select the types of uses (select one or both, as applicable). | ☑ Prescription Use (Part 21 CFR 801 Subpart D) ☐ Over-The-Counter Use (21 CFR 801 Subpart C) | ? | Page 1 of 1 {4} K260008 510(k) Summary # 510(k) SUMMARY ## Deterioration Index Version 2 Verona, WI 53593 Owner: Epic Systems Corporation Contact Person: Brian Jacobson Phone: 1 (608) 271-9000 Fax: 1 (608) 271-7237 Date Prepared: 09/17/2026 **Trade Name:** Deterioration Index Version 2 **Classification Regulation:** 21 CFR 870.2210 – Adjunctive predictive cardiovascular indicator **Classification:** Class II **Product Code:** QNL **510(k) Number:** K260008 **Predicate Device:** eCARTv5 Clinical Deterioration Suite (“eCART”) (K233253), (product code: QNL) ## 1. Device Description Deterioration Index Version 2 (DIV2) is clinical decision support software that is used to help identify patients at risk of clinical deterioration. A patient is at risk of clinical deterioration if one of the following two outcomes is predicted: (1) escalation of care in the next 24 hours or (2) mortality in the next 72 hours. An escalation of care is a transfer to the ICU, a rapid response team call, or a code. DIV2 is for use among patients 18 years of age and older admitted to the hospital in an inpatient unit, observation unit, or on observation status in the emergency department of a hospital. DIV2 is not intended to be used in settings including long-term care, the operating room, a post anesthesia care unit (PACU), labor and delivery, or for non-observation patients in an emergency department. DIV2’s end users are trained healthcare professionals. DIV2 must be used with the Epic electronic health record (EHR) system. DIV2 uses documentation of data received directly from the EHR, such as routine vital signs, laboratory results, nursing assessments, medication administrations, patient procedures, Page 1 of 10 {5} K260008 510(k) Summary healthcare utilization information, diagnoses, and patient demographics. DIV2 derives values such as averages, standard deviation, minimum recent value, maximum recent value, or the change between the previous two values (slope) for some vitals and laboratory results. DIV2 uses 195 inputs to calculate a risk score that can be used to identify patients at risk of clinical deterioration. The score is calculated by a static gradient boosted random forest (trained with LightGBM) algorithm. DIV2's score is scaled from 0-100 based on the positivity (flag rate). Trained healthcare professionals at an organization determine the scores at which further patient evaluation should occur. The score is sent to the EHR, where it can be displayed in a user interface and provide information to trained clinicians. This enables clinicians to follow organization-defined clinical policies for assessing deterioration risk and follow the corresponding workflows for at-risk patients. DIV2 includes a "global model". The global model is a static gradient boosted random forest (trained with LightGBM). ## 2. Intended Use / Indications for Use ### 2.1. Intended Use Deterioration Index Version 2 is intended to provide trained healthcare professionals with a patient score that reflects the underlying patient condition to supplement informed decision-making. ### 2.2. Indications for Use Deterioration Index Version 2 is software that calculates the risk of patient deterioration, defined as the patient experiencing an escalation of care (a transfer to the ICU, a rapid response team call, or a code) or mortality. Deterioration Index Version 2 is for use with adult patients aged 18 or older who are admitted to an inpatient unit, observation unit, or on observation status in the emergency department of a hospital. The software is not for use with patients in a long-term care setting, operating room, post anesthesia care unit (PACU), labor and delivery unit, or for non-observation patients in an emergency department. The software is for use by trained healthcare professionals. Deterioration Index Version 2's risk score is calculated from Epic EHR data, such as vitals that have been documented for the patient. As clinical decision support software, Deterioration Index Version 2's risk score is intended to aid clinicians in identifying which patients are more likely to clinically deteriorate. Deterioration Index Version 2 provides additional information and does not replace the standard of care or clinical judgment. The device is not intended to serve as the primary determinant of health status for a patient, nor as a substitute for a healthcare professional's judgment and decision making. The device is not a primary diagnostic or treatment tool. Information provided by the device is not intended to be relied upon as the sole basis for decisions that have an impact that may cause death or an irreversible deterioration of a person's state of health. ## 3. Comparison with the Predicate Device Deterioration Index Version 2 and the predicate device are similar in producing a risk-predictive output using a machine-learning algorithm. Inputs to the software device include vital signs, assessments, and laboratory Page 2 of 10 {6} K260008 510(k) Summary data collected from the hospital EHR system. Both produce a real-time output of a score representing the potential risk of patient deterioration for hospitalized adult patients. The score is filed to the EHR so that clinical professionals can incorporate the score into their EHR workflows, applying their independent professional medical judgment as to if and how they incorporate the score into their practice. The differences in indications for use from the predicate device are not critical to the intended use of DIV2, nor do they raise different questions of safety or effectiveness when the subject device is used as labeled. ### 3.1. Differences in Technological Characteristics Compared with the predicate device, DIV2 predicts a broader composite outcome (including rapid response team calls and code events), uses a longer mortality prediction horizon (72 vs. 24 hours), and generates scores on a scheduled cadence rather than being triggered by vital sign documentation. DIV2 was also trained on a broader patient population and incorporates additional inputs, such as admission information, medications, LDAs, and diagnoses, but these differences do not alter its intended use or raise new safety or effectiveness questions. With respect to the output, the predicate device (eCARTv5)'s risk score was validated on the outcome of death or ICU transfer and included training and testing on data from adult ward patients. Performance data for DIV2, including retrospective testing, demonstrates that the subject device is substantially equivalent to the predicate device. Software development and testing were conducted in accordance with EN IEC 62304:2006+A1:2015 Medical device software – Software life-cycle processes. ## 4. Performance Summary ### 4.1. Model validation testing A retrospective model validation study was conducted to demonstrate that Deterioration Index Version 2 provides acceptable performance using data from three geographically different sites in the United States. The three sites are each one or more of the following: academic hospitals, critical access facilities, research centers, and safety-net organizations. There are multiple methods to statistically validate a predictive model in an acute setting. This performance summary includes the results calculated using two of the methodologies from the study, for easier comparison with other risk scores: Methodology 1 - uses a combined target of an escalation including ICU stay, a rapid response team call, or a code event within 24 hours, or mortality in 72 hours for hospitalized patients or patients on observation status in the ED, aged 18 and older; bootstrapping aggregation Methodology 2 - uses a combined target of an escalation including ICU stay or mortality in 24 hours for hospitalized patients, excluding observation patients, adults; bootstrapping aggregation ### 4.2. Primary Endpoint The primary endpoint used in methodology 1 to assess the following performance metrics for the risk level generated by Deterioration Index Version 2 model with 3 risk levels from 2 thresholds: - Monotonic increase in deterioration stratum-specific likelihood ratio (SSLR) as risk level severity increases Page 3 of 10 {7} K260008 510(k) Summary • Non-overlapping SSLRs in each of the adjacent risk levels, along with 95% confidence intervals (CIs) The following table shows a monotonic increase in SSLR as the risk level increases while also showing non-overlapping SSLRs in adjacent risk levels. Outcome Prevalence: 3.8% (410,520/10,724,900) | Risk Level | n - observations | SSLR (95% CI) | | --- | --- | --- | | Low (0-70) | 9,735,278 | 0.55 (0.55-0.56) | | Medium (70-90) | 657,844 | 3.78 (3.60-3.96) | | High (90-100) | 331,778 | 13.21 (12.69-13.80) | ### 4.3. Secondary Endpoints The secondary endpoints include the following performance metrics, including 2-sided 95% CIs for all statistics: - Area Under the Receiver Operating Characteristic (ROC) curve (i.e., AUROC, C-Statistic) - A three-tier confusion matrix separating each risk level, and a two-tier confusion matrix showing statistics using either the medium or high risk threshold on their own. #### Area Under the Receiver Operating Characteristic Curve (C-Statistic) The C-statistic and 95% confidence interval for the full validation set using methodology 1 is 0.83 (0.83-0.84) when using DIV2's combined target, and 0.94 (0.94-0.94) when using only mortality as the target. The C-statistic and 95% confidence interval for the full validation set using methodology 2 is 0.81 (0.81-0.82) when using the combined target, and 0.95 (0.94-0.95) when using only mortality as the target. #### Confusion Matrices ##### Methodology 1 A full three-tier confusion matrix calculated using methodology 1 is shown below. | Risk Level | Total Positive | Total Negative | | --- | --- | --- | | Low (0-70) | 210,265 | 9,525,013 | | Medium (70-90) | 85,930 | 571,914 | | High (90-100) | 114,325 | 217,453 | A two-tier confusion matrix calculated using methodology 1 for the medium and high-risk thresholds separately is shown below. Combined target prevalence: 3.8% | Threshold | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | Flag rate (95% CI) | | --- | --- | --- | --- | --- | | Medium (≥70) | 0.49 (0.48-0.50) | 0.92 (0.92-0.92) | 0.20 (0.20-0.21) | 0.09 (0.09-0.09) | Page 4 of 10 {8} K260008 510(k) Summary | Threshold | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | Flag rate (95% CI) | | --- | --- | --- | --- | --- | | High (≥90) | 0.28 (0.27-0.29) | 0.98 (0.98-0.98) | 0.34 (0.33-0.35) | 0.03 (0.03-0.03) | An additional confusion matrix similar to the previous but calculated using only mortality as the ground truth label is shown below. Mortality prevalence: 1.4% | Threshold | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | Flag rate (95% CI) | | --- | --- | --- | --- | --- | | Medium (≥70) | 0.80 (0.80-0.81) | 0.92 (0.92-0.92) | 0.12 (0.12-0.12) | 0.09 (0.09-0.09) | | High (≥90) | 0.56 (0.55-0.57) | 0.98 (0.98-0.98) | 0.25 (0.24-0.26) | 0.03 (0.03-0.03) | ## Methodology 2 A three-tier confusion matrix calculated using methodology 2 is shown below. | Risk Level | Total Positive | Total Negative | | --- | --- | --- | | Low (0-70) | 246,753 | 7,761,088 | | Medium (70-90) | 96,076 | 537,210 | | High (90-100) | 107,244 | 226,029 | A two-tier confusion matrix calculated using methodology 2 for the medium and high-risk thresholds separately is shown below. Combined target prevalence: 5.0% | Threshold | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | Flag rate (95% CI) | | --- | --- | --- | --- | --- | | Medium (≥70) | 0.45 (0.44-0.46) | 0.91 (0.91-0.91) | 0.21 (0.21-0.21) | 0.11 (0.11-0.11) | | High (≥90) | 0.24 (0.23-0.24) | 0.97 (0.97-0.97) | 0.32 (0.31-0.33) | 0.04 (0.04-0.04) | An additional confusion matrix similar to the previous but calculated using only mortality as the ground truth label is shown below. Mortality prevalence: 1.0% | Threshold | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | Flag rate (95% CI) | | --- | --- | --- | --- | --- | | Medium (≥70) | 0.85 (0.83-0.87) | 0.90 (0.90-0.90) | 0.08 (0.07-0.08) | 0.11 (0.11-0.11) | | High (≥90) | 0.62 (0.60-0.64) | 0.97 (0.97-0.97) | 0.16 (0.16-0.17) | 0.04 (0.04-0.04) | Page 5 of 10 {9} K260008 510(k) Summary ### 4.4. Subgroup Analysis Results are also reported for the following subgroups: Sex: Male vs. Female | Methodology 1 | | | | | Methodology 2 | | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | | Sex | Prevalence | AUROC (95% CI) | Risk Level | SSLR (95% CI) | Prevalence | AUROC (95% CI) | Risk Level | SSLR (95% CI) | | **Male** | 4.4% (209,179/4,786,900) | 0.81 (0.81-0.82) | Low (0-70) | 0.58 (0.56-0.59) | 5.7% (231,511/4,060,700) | 0.79 (0.79-0.80) | Low (0-70) | 0.62 (0.61-0.64) | | | | | Medium (70-90) | 3.17 (3.00-3.35) | | | Medium (70-90) | 2.95 (2.75-3.14) | | | | | High (90-100) | 10.83 (10.13-11.52) | | | High (90-100) | 7.41 (7.01-7.86) | | **Female** | 3.4% (200,003/5,899,500) | 0.85 (0.84-0.85) | Low (0-70) | 0.53 (0.52-0.55) | 4.4% (217,077/4,882,700) | 0.83 (0.83-0.83) | Low (0-70) | 0.58 (0.57-0.59) | | | | | Medium (70-90) | 4.43 (4.19-4.73) | | | Medium (70-90) | 3.85 (3.63-4.06) | | | | | High (90-100) | 15.94 (15.04-16.96) | | | High (90-100) | 10.82 (10.28-11.41) | | Methodology 1 | | | | | | Methodology 2 | | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | | Sex | Threshold | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | Flag rate (95% CI) | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | Flag rate (95% CI) | | **Male** | Medium (≥70) | 0.47 (0.46-0.48) | 0.91 (0.91-0.91) | 0.20 (0.19-0.20) | 0.10 (0.10-0.11) | 0.44 (0.43-0.45) | 0.90 (0.90-0.90) | 0.21 (0.20-0.21) | 0.12 (0.12-0.12) | | | High (≥90) | 0.28 (0.27-0.29) | 0.97 (0.97-0.98) | 0.33 (0.32-0.34) | 0.04 (0.04-0.04) | 0.23 (0.22-0.24) | 0.97 (0.97-0.97) | 0.31 (0.30-0.32) | 0.04 (0.04-0.04) | | **Female** | Medium (≥70) | 0.50 (0.49-0.51) | 0.93 (0.93-0.93) | 0.21 (0.20-0.21) | 0.08 (0.08-0.08) | 0.46 (0.45-0.47) | 0.92 (0.92-0.92) | 0.21 (0.21-0.22) | 0.10 (0.10-0.10) | | | High (≥90) | 0.28 (0.27-0.29) | 0.98 (0.98-0.98) | 0.36 (0.34-0.37) | 0.03 (0.03-0.03) | 0.24 (0.23-0.25) | 0.98 (0.98-0.98) | 0.33 (0.32-0.35) | 0.03 (0.03-0.03) | Page 6 of 10 {10} K260008 510(k) Summary Race: White vs. Black vs. Other/Unknown | Methodology 1 | | | | | Methodology 2 | | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | | Race | Prevalence | AUROC (95% CI) | Risk Level | SSLR (95% CI) | Prevalence | AUROC (95% CI) | Risk Level | SSLR (95% CI) | | **White** | 4.1% (350,388/ 8,487,600) | 0.83 (0.83-0.84) | Low (0-70) | 0.53 (0.52-0.54) | 4.9% (350,014/ 7,208,300) | 0.81 (0.81-0.82) | Low (0-70) | 0.57 (0.56-0.58) | | | | | Medium (70-90) | 3.68 (3.50-3.85) | | | Medium (70-90) | 3.26 (3.10-3.41) | | | | | High (90-100) | 12.89 (12.34- 13.46) | | | High (90-100) | 9.05 (8.66-9.46) | | **Black** | 2.9% (34,277/ 1,172,200) | 0.79 (0.78-0.81) | Low (0-70) | 0.77 (0.74-0.80) | 6.6% (58,543/883,900) | 0.83 (0.82-0.83) | Low (0-70) | 0.73 (0.71-0.75) | | | | | Medium (70-90) | 3.41 (2.76-4.07) | | | Medium (70-90) | 4.51 (3.96-5.23) | | | | | High (90-100) | 11.63 (9.15- 13.92) | | | High (90-100) | 9.70 (8.31-11.54) | | **Other/ Unknown** | 2.4% (25,855/ 1,065,100) | 0.83 (0.82-0.85) | Low (0-70) | 0.61 (0.57-0.64) | 4.7% (41,516/882,200) | 0.82 (0.81-0.83) | Low (0-70) | 0.66 (0.64-0.68) | | | | | Medium (70-90) | 4.42 (3.71-5.13) | | | Medium (70-90) | 4.14 (3.67-4.85) | | | | | High (90-100) | 13.72 (11.35- 16.59) | | | High (90-100) | 9.92 (8.26-11.95) | | Methodology 1 | | | | | | Methodology 2 | | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | | Race | Threshold | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | Flag rate (95% CI) | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | Flag rate (95% CI) | | **White** | Medium (≥70) | 0.51 (0.51- 0.52) | 0.92 (0.92- 0.92) | 0.21 (0.21- 0.22) | 0.10 (0.10- 0.10) | 0.48 (0.47- 0.49) | 0.90 (0.90- 0.90) | 0.20 (0.20-0.21) | 0.12 (0.11-0.12) | | | High (≥90) | 0.30 (0.29- 0.30) | 0.98 (0.98- 0.98) | 0.36 (0.34- 0.37) | 0.03 (0.03- 0.04) | 0.26 (0.26- 0.27) | 0.97 (0.97- 0.97) | 0.32 (0.31-0.33) | 0.04 (0.04-0.04) | | **Black** | Medium (≥70) | 0.26 (0.24- 0.30) | 0.95 (0.95- 0.95) | 0.14 (0.12- 0.15) | 0.06 (0.05- 0.06) | 0.31 (0.29- 0.33) | 0.95 (0.94- 0.95) | 0.29 (0.27-0.32) | 0.07 (0.07-0.07) | | | High (≥90) | 0.14 (0.11- 0.16) | 0.99 (0.99- 0.99) | 0.26 (0.22- 0.30) | 0.02 (0.01- 0.02) | 0.12 (0.11- 0.14) | 0.99 (0.99- 0.99) | 0.41 (0.37-0.45) | 0.02 (0.02-0.02) | Page 7 of 10 {11} K260008 510(k) Summary | Methodology 1 | | | | | | Methodology 2 | | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | | Race | Threshold | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | Flag rate (95% CI) | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | Flag rate (95% CI) | | **Other/Unknown** | Medium (≥70) | 0.43 (0.40-0.47) | 0.94 (0.93-0.94) | 0.15 (0.13-0.16) | 0.07 (0.07-0.07) | 0.39 (0.36-0.41) | 0.93 (0.93-0.94) | 0.22 (0.20-0.24) | 0.08 (0.08-0.09) | | | High (≥90) | 0.23 (0.20-0.26) | 0.98 (0.98-0.99) | 0.25 (0.22-0.29) | 0.02 (0.02-0.02) | 0.18 (0.16-0.20) | 0.98 (0.98-0.98) | 0.33 (0.29-0.38) | 0.03 (0.02-0.03) | The statistics shown in the table below are from a block aggregation analysis for the medium-risk threshold (≥ 70). Cohorts where fewer than 100 escalation events or mortalities were observed are italicized in the table below. | | Race | Blocks (% total) | Prevalence within cohort | Sensitivity (95% CI) | PPV (95% CI) | Flag Rate (95% CI) | | --- | --- | --- | --- | --- | --- | --- | | **Site 1** | White | 77,983 (95.6%) | 9.8% | 0.62 (0.61, 0.63) | 0.19 (0.19, 0.20) | 0.32 (0.31, 0.32) | | | Black | 1,795 (2.2%) | 10.1% | 0.63 (0.56, 0.70) | 0.22 (0.19, 0.26) | 0.29 (0.27, 0.31) | | | Other/ Unknown | 1,791 (2.2%) | 8.4% | 0.61 (0.53, 0.69) | 0.19 (0.16, 0.23) | 0.27 (0.25, 0.29) | | **Site 2** | White | 14,764 (61.5%) | 4.7% | 0.54 (0.51, 0.58) | 0.15 (0.14, 0.16) | 0.17 (0.17, 0.18) | | | Black | 1,560 (6.5%) | 4.2% | 0.45 (0.33, 0.57) | 0.12 (0.09, 0.17) | 0.15 (0.13, 0.17) | | | Other/ Unknown | 7,671 (32.0%) | 4.9% | 0.58 (0.52, 0.62) | 0.15 (0.13, 0.17) | 0.19 (0.18, 0.20) | | **Site 3** | White | 5,028 (27.9%) | 7.5% | 0.34 (0.30, 0.39) | 0.16 (0.14, 0.19) | 0.16 (0.15, 0.17) | | | Black | 9,631 (53.5%) | 6.1% | 0.32 (0.28, 0.35) | 0.15 (0.13, 0.17) | 0.13 (0.12, 0.14) | | | Other/ Unknown | 3,336 (18.5%) | 6.5% | 0.44 (0.38, 0.51) | 0.20 (0.17, 0.24) | 0.15 (0.13, 0.16) | Page 8 of 10 {12} K260008 510(k) Summary Ethnicity: Hispanic vs. Not Hispanic | Methodology 1 | | | | | Methodology 2 | | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | | Ethnicity | Prevalence | AUROC (95% CI) | Risk Level | SSLR (95% CI) | Prevalence | AUROC (95% CI) | Risk Level | SSLR (95% CI) | | Hispanic | 1.7% (25,433/1,462,600) | 0.81 (0.80-0.83) | Low (0-70) | 0.72 (0.69-0.75) | 4.5% (54,179/1,209,100) | 0.83 (0.82-0.83) | Low (0-70) | 0.69 (0.67-0.71) | | | | | Medium (70-90) | 4.05 (3.24-4.92) | | | Medium (70-90) | 4.55 (4.00-5.08) | | | | | High (90-100) | 10.98 (9.04-13.18) | | | High (90-100) | 10.99 (9.37-12.82) | | Not Hispanic | 4.2% (385,087/9,262,300) | 0.83 (0.83-0.83) | Low (0-70) | 0.54 (0.54-0.55) | 5.1% (395,894/7,765,300) | 0.81 (0.81-0.81) | Low (0-70) | 0.59 (0.58-0.60) | | | | | Medium (70-90) | 3.67 (3.51-3.83) | | | Medium (70-90) | 3.27 (3.10-3.41) | | | | | High (90-100) | 12.92 (12.40-13.51) | | | High (90-100) | 8.77 (8.39-9.15) | | Methodology 1 | | | | | | Methodology 2 | | | | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | | Ethnicity | Threshold | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | Flag rate (95% CI) | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | Flag rate (95% CI) | | Hispanic | Medium (≥70) | 0.32 (0.29-0.35) | 0.95 (0.94-0.95) | 0.09 (0.08-0.11) | 0.06 (0.06-0.06) | 0.35 (0.33-0.37) | 0.94 (0.94-0.95) | 0.22 (0.21-0.24) | 0.07 (0.07-0.07) | | | High (≥90) | 0.16 (0.13-0.18) | 0.99 (0.98-0.99) | 0.16 (0.13-0.20) | 0.02 (0.02-0.02) | 0.15 (0.14-0.17) | 0.99 (0.98-0.99) | 0.34 (0.31-0.38) | 0.02 (0.02-0.02) | | Not Hispanic | Medium (≥70) | 0.50 (0.49-0.51) | 0.92 (0.92-0.92) | 0.21 (0.21-0.22) | 0.10 (0.10-0.10) | 0.47 (0.46-0.47) | 0.91 (0.90-0.91) | 0.21 (0.20-0.21) | 0.11 (0.11-0.12) | | | High (≥90) | 0.29 (0.28-0.29) | 0.98 (0.98-0.98) | 0.36 (0.35-0.37) | 0.03 (0.03-0.03) | 0.25 (0.24-0.26) | 0.97 (0.97-0.97) | 0.32 (0.31-0.33) | 0.04 (0.04-0.04) | Page 9 of 10 {13} K260008 510(k) Summary ## 5. Conclusion The differences between the predicate and the subject device do not raise any new or different questions of safety or effectiveness. The information and testing presented in this 510(k) Summary demonstrate that the Deterioration Index Version 2 is substantially equivalent to the predicate device cleared under K233253. Page 10 of 10