Invistics' Pivot to Hospital Drug Diversion Detection

The case study focuses on Tom Knight, the founder and CEO of Invistics Corporation, a company that initially specialized in inventory management software for manufacturing, particularly pharmaceutical manufacturing. The central dilemma presented is whether Invistics should pivot and enter the hospital market with a drug diversion detection technology.

Tom Knight's Background and Invistics' Origins

Tom Knight has an MIT background, holding a master's degree in management science and engineering and a bachelor's in mechanical engineering. Before founding Invistics, he worked for Alcoa, a global industrial corporation in aluminum extraction and manufacturing, and then Siemens. It was at Siemens that the idea for Invistics sparked.

Invistics began by addressing inventory management in manufacturing, a field undergoing a "just-in-time" revolution. Manufacturers sought efficiency and fine-grained understanding of components. Tom Knight recognized the potential for software to manage this complex process.

The Problem of Drug Diversion in Hospitals

The discussion then shifted to the problem of drug diversion in hospital settings. Drug diversion refers to the illicit use or theft of prescription medications, particularly controlled substances like opioids.

Examples of Drug Diversion

Attendees shared various examples of how drug diversion occurs:

• Oral Morphine Theft: Oral morphine, often in large bottles, is sometimes left on drug trolleys manned by a single nurse. Discrepancies between prescribed amounts and remaining quantities have been observed, sometimes correlating with specific nurses or locations.
• Sophisticated Methods: As detection methods become more advanced, so do diversion techniques.
  • Partial Dosing: A controlled substance is prescribed, checked out by a nurse, but only a portion (or none) is given to the patient, with the remainder pocketed by the practitioner.
  • Vial Tampering: Vials are drilled with tiny holes, contents drained and replaced with water, then returned to the shelf.
  • Manipulation of Records: Pill counts can be falsified.
  • Theft from Supply Chain: Drugs can disappear from loading docks or during the "wasting" process (where excess medication is supposed to be discarded but is instead diverted).

Impact of Drug Diversion

Drug diversion has significant consequences for various stakeholders:

• Patients: Patients may not receive the pain medication they need, leading to continued pain and potentially being mislabeled as "drug-seeking." This creates a vicious cycle.
• Diverters: The individuals diverting drugs are often victims themselves, developing substance use disorders. Early detection can lead to earlier intervention and a higher chance of recovery.
• Institutions (Hospitals):
  • Legal and Financial Penalties: Hospitals face massive fines (multimillion-dollar cases are common) from regulatory bodies like the DEA for poor control over controlled substances.
  • Liability: Hospitals can be sued by patients harmed by diversion (e.g., receiving saline instead of fentanyl during a procedure).
  • Reputational Damage: A single diversion incident can tarnish the reputation of an entire department or hospital.
  • Supply Issues: Regulatory controls on distributors can lead to supply chain disruptions for legitimate patient care.
• Society: Diversion contributes to the broader opioid epidemic, with diverted drugs often ending up on the street, sometimes as illicit fentanyl.

Challenges in Addressing Diversion

• Reporting Concerns: Colleagues may be aware of diversion but lack a safe, anonymous mechanism to report it without immediately jeopardizing the diverter's career or creating personal conflict.
• Defining "Poor Control": The definition of "poor control" by regulatory bodies can be ambiguous, making it unclear whether preventative or investigative measures are sufficient.

The Decision: Enter the Hospital Market?

Tom Knight and Invistics faced a critical decision: should they enter the hospital market with their drug diversion technology?

Arguments for Entering the Market

• Unmet Need: There is a tremendous, widespread, and recognized unmet need for effective drug diversion detection.
• Existing Infrastructure: Hospitals already have staff dedicated to managing controlled substances, indicating a pre-existing "cognizant agency" and budget for this area.
• Regulatory Pressure & Liability: Hospitals are highly regulated and face significant financial penalties and legal liability for diversion, making risk mitigation a strong motivator.
• Societal Momentum: There is growing societal awareness, attention, and potential funding opportunities related to the opioid crisis.
• Technology Applicability: Invistics' core technology in inventory control and supply chain management is adaptable. While drug diversion detection requires different algorithms, the underlying logic can be applied.
• Superior Solution: Invistics' technology is known to be superior to existing solutions, offering more accurate and earlier detection.
• Revenue Generation: Beyond risk mitigation, the software can improve "charge capture" by ensuring all administered medications are properly billed, potentially generating significant revenue for hospitals.
• NIH Backing: The availability of NIH Small Business Innovation Research (SBIR) grants provided crucial funding and credibility, allowing Invistics to build a specialized team and approach hospitals with a research-backed offering.

Arguments Against Entering the Market

• Market Differences: The hospital market is vastly different from manufacturing.
  • Sales Force: Invistics' sales force was attuned to manufacturing; a new sales force with healthcare expertise would be needed, incurring significant cost and time.
  • Software Adaptation: While similarities exist, the software for drug diversion detection is "totally different" from manufacturing inventory management, requiring new algorithms and development.
  • Implementation Complexity: Hospital decision-making processes are lengthy and complicated, involving multiple stakeholders (pharmacy, ethics committees, IT). Implementation requires integrating with disparate systems (EMR, automated dispensing cabinets), leading to high costs and extended timelines.
• Training Data: If the existing machine learning models were not trained on relevant drug diversion cases within healthcare systems, their predictive accuracy could be compromised.
• Market Resistance: Hospitals might be resistant to "disruptive" or "too innovative" products, especially if they challenge existing habits or require significant workflow changes.
• "First Mover" Disadvantage: Hospitals are often reluctant to be the first to adopt a new technology, preferring to see it proven elsewhere.

Tom Knight's Decision and Outcome

Despite the challenges, Tom Knight decided to enter the hospital market. His primary motivations were:

• Unmet Need and Victims: The profound impact of drug diversion on patients, diverters, and institutions, coupled with the potential to save lives and aid recovery, was a strong driver.
• NIH Funding: The SBIR grant from NIDA provided the necessary capital and credibility to build a new team with clinical and health backgrounds and to approach hospitals with an NIH-funded study. This allowed Invistics to frame their offering as participation in a research study rather than just a software sale, which was more appealing to some hospitals.
• Team Motivation: Many team members were personally affected by the opioid epidemic and were motivated by the opportunity to make a significant impact.

Key Learnings and Strategies

• Targeting Hospitals: Invistics initially targeted hospitals with mature drug diversion prevention programs that were already spending significant amounts on detection. They also found that large academic medical centers were particularly receptive.
• Data Collection and Model Training: They focused on acute care inpatient hospitals and initially on nursing data, as nurses were numerous and provided ample training data. They worked with hospitals that had existing diversion cases to train their models.
• Expanding Data Sources: To improve detection, Invistics expanded data sources beyond EMR and automated dispensing cabinets to include employee time clock records, purchasing records from distributors, and central pharmacy data. This helped detect different types of diversion, including large-scale theft for resale.
• Pricing Model: The pricing model shifted from factory size to the number of hospital beds, with annual subscriptions and quantity discounts for integrated delivery networks.
• Addressing Data Interoperability: They focused on understanding major EMR (Epic, Cerner) and automated dispensing cabinet (Pyxis, Omnicell) systems, acknowledging that even within these, customization required significant effort.
• Intellectual Property: Licensing agreements ensured Invistics could use customer data (anonymized and consolidated) to continuously train and improve their models.
• Community Building: Invistics established a customer advisory board and held "grand rounds" where hospitals shared new diversion patterns, fostering collective learning.
• Non-Profit Initiative: They co-founded healthcareddiversion.org, a non-profit, to engage a broader range of stakeholders (state boards of health, medical boards, pharmacy boards, CDC) and promote education and support for recovery.
• Human-Centered Approach: Tom Knight emphasized shifting the narrative from viewing diverters as criminals to recognizing them as victims of substance use disorder, aiming for early detection to facilitate recovery. He stressed the importance of professional and discreet investigations to avoid false accusations.
• Pivoting: Tom Knight highlighted the importance of being willing to pivot when market signals indicate a better opportunity, even if it means reallocating resources from a successful existing market.

Outcome

Invistics successfully entered the hospital market, growing to serve nearly 200 hospitals, including many large academic medical centers. They were eventually acquired by Wolters Kluwer in June 2023. This acquisition provided significant financial rewards and, crucially, allowed for a massive expansion of the team and market penetration, leveraging Wolters Kluwer's existing relationships with 6,000 hospitals. The software consistently ranked as the best in the market, and its lead continues to grow.

Tom Knight also shared his current venture, Optinosis, which aims to use AI and large clinical datasets to detect cancer earlier, demonstrating his continued belief in AI's potential for good in healthcare.