Growing Caseloads and Complex Investigations Are Testing Licensing Board Systems

Complaint growth can change the workload beneath the headline

The California Board of Registered Nursing’s November 2025 enforcement materials illustrate the importance of looking inside an intake total. Public complaints increased from 3,682 in fiscal year 2021–22 to 5,330 in 2024–25, a rise of approximately 44.8%. Across all intake categories, however, the total increased from 7,739 to 8,317, approximately 7.5%. Applicant-related intake declined over the same period. The board’s public-complaint category also includes referrals from other state agencies; it is broader than complaints submitted by individual consumers. [1]

A modest change in the combined total can therefore coexist with a substantial change in the kinds of matters reaching staff. Those figures do not tell us how many hours each category requires. They do demonstrate why a single aggregate number cannot adequately describe the operational challenge.

The board’s FY 2026–27 budget proposal provides a second perspective. It reports that complaints referred to its Investigations Division grew 64% since 2020–21 and that pending, unassigned investigations increased 193% between 2020–21 and 2024–25. The proposal requested eight additional special investigators and $1.43 million for the first fiscal year. Those are figures from a staffing request, not evidence that the requested resources were ultimately approved. [2]

An unassigned investigation is a different operational problem from an investigation awaiting outside records or a matter ready for a hearing. Each may appear in an open-case count, but each requires a different intervention. Hiring capacity, better assignment practices, timely records production, and hearing availability address different constraints. A useful management system must make those distinctions visible.

Aging cases expose the limits of aggregate performance measures

Arizona’s September 2025 nursing-board audit found that 56% of complaints closed between July 2024 and January 2025 had not been resolved within 180 days. As of January 2025, the board had 2,177 open complaints; 1,543 had been unresolved for more than 180 days, including 51 classified as high risk. These are historical audit findings, not a count of the board’s current inventory. [3]

The accompanying exchange between the board and auditors is particularly instructive. The board cited growth from 3,067 complaints in 2011 to 5,378 in 2024 and sought more staff. Auditors cautioned that the board’s approximately 5,000 annual matters included two different categories: complaint investigations and licensing investigations. The latter generally had much shorter actual processing times and a different applicable timeframe. The auditors also reported that the board agreed with the findings and planned to implement the recommendations. [4]

This distinction matters for public discussion as well as internal management. Combining unlike matters can obscure the experience of the cases that need sustained investigative attention. It can also make an agency look more or less efficient depending on how its workload is categorized.

A board should be able to see how long a case has been open and how long it has occupied its current stage. It should also be able to distinguish a missing record from an uncompleted internal task. Without that context, a report of aging cases identifies a symptom while leaving staff to reconstruct its cause.

Complexity is a separate source of workload

Case volume counts matters. It does not capture how many people, documents, disputed events, or outside organizations each matter involves.

California’s nursing-board budget proposal identifies medical spas and intravenous hydration services as sources of complex investigations involving multiple subjects and additional coordination. It also acknowledges that specific complaint tracking codes for these services had not yet been established. The board’s description supports the conclusion that these practice settings create difficult investigative work; it does not provide a reliable percentage of all complaints attributable to them. [2]

Credentialing investigations offer another example. In its 2025 annual report, the Texas Board of Nursing reported 367 open cases at fiscal year-end involving further investigation of educational qualifications from 11 schools in connection with Operation Nightingale. It reported 121 cases resolved through disciplinary action, denial of initial licensure, or license inactivation. Those outcomes should not be collapsed into a single count of disciplinary findings, and an open investigation should not be treated as proof of wrongdoing. [5]

When many cases arise from a common institution or event, organizing the shared background is only part of the task. Each person’s record still requires individual evaluation. A system should allow authorized staff to connect related matters without silently importing an allegation or conclusion from one person’s case into another’s.

New Mexico’s Board of Nursing describes the routine sources of investigative delay in practical terms: investigator workload, the complexity of the issues, the volume of documentation, obtaining outside records, and competing priority cases. It also explains that a complaint notification is an allegation under review, rather than a determination that a violation occurred. [6]

Together, these examples suggest a better way to understand capacity. The relevant question is how much work a case requires to reach a supported decision, including the time spent locating, reconciling, and checking information. A count of files alone cannot answer it.

New models of practice bring new evidentiary questions

The regulatory questions themselves are also changing. In May 2026, the Federation of State Medical Boards announced a workgroup on regulating artificial intelligence in medical practice, with particular attention to tools performing clinical functions with limited or no direct physician supervision. Its remit includes oversight responsibilities, documentation, consent, and standards of care. This is evidence of an expanding regulatory agenda, not evidence that AI-related complaints have already increased by a measurable amount. [7]

For a future investigation involving such a tool, a reviewer may need to understand what information the clinician saw, what the software generated, when the output changed, and who exercised professional judgment. That hypothetical example shows why preserving the source and sequence of evidence matters. A narrative summary without a traceable underlying record may omit the very distinction that determines responsibility.

The same design principle applies to more familiar disputes. A chronology should distinguish when an event happened from when it was reported. A document should retain its origin and version. A reviewer’s interpretation should remain distinguishable from a witness’s statement and from the board’s eventual finding.

Fragmented case administration creates avoidable work

Virginia’s September 2025 audit of the Department of Professional and Occupational Regulation offers a direct connection between case organization and agency performance. The inspector general found that different tracking systems impeded comprehensive monitoring from intake through resolution. Approximately 300 complaints awaited intake review on April 29, 2025. [8]

The audit also identified 314 of 906 reviewed open cases as remaining open more than 150 days. Many were no longer in the investigative stage covered by the agency’s 107-day target, so that statistic cannot be read as 314 missed investigative deadlines. Twenty-nine of those 314 cases were delayed by data-entry errors. The findings point to the importance of accurate stage information as well as elapsed time. [8]

An agency needs a reliable account of what has happened and what should happen next. If a file moves to legal review while a separate tracker still labels it an investigation, managers may direct attention to the wrong queue. If a requested document arrives without being connected to the outstanding request, staff may repeat work that is already complete.

These are examples of failures that careful workflow design should help prevent. Technology cannot guarantee that a witness responds, create an expert’s availability, or eliminate contested facts. It can make those dependencies explicit and help people distinguish necessary waiting from an overlooked task.

What better systems for active cases should provide

The research points toward a practical design agenda. Boards need a coherent working record that supports professional judgment throughout a matter. The following are proposed requirements, rather than claims about any particular product’s deployed capabilities.

First, the case record should connect allegations to the documents and responses relevant to them. Reviewers need to trace a summary back to its source, inspect contrary evidence, and see what remains unresolved. Organizing information must preserve disagreement rather than flatten it into a premature conclusion.

Second, responsibility and timing should be explicit. An active matter should show its current owner, the next required action, the applicable deadline, and any dependency preventing progress. Deadline calculations should be reviewable and configurable to the agency’s actual procedures. Exceptions and extensions should remain visible in the history.

Third, access should follow a person’s authorized role in the particular matter. Investigators, counsel, board members, complainants, and respondents do not necessarily have the same rights to information. Record retention, disclosure, redaction, and audit history need to be designed around the agency’s requirements. Where a provider serves both professionals and regulators, separation of confidential information requires particular care.

Fourth, assistance with document organization or summarization should remain subject to human verification. If AI is used, the system should identify its output, preserve source references, and allow corrections. Decisions about credibility, legal sufficiency, public risk, and discipline require accountable human review under the applicable process.

Finally, a board should be able to evaluate whether a new system actually helps. A pilot could measure time spent preparing a case for review, repeated requests for the same material, missing-document rates, and aging within individual stages. Those measures should be paired with checks on record completeness, errors, and the ability to correct them. Faster closure by itself does not demonstrate a better process.

These recommendations require operational work as well as software. Agencies must define their case stages, agree on ownership, validate migrated records, and train the people who use the system. Any evaluation should account for staffing changes and the mix of cases before attributing improvements to technology.

BoardWise is beginning this work with boards and agencies

BoardWise has begun working with licensing boards and agencies on systems to help better organize and manage active cases. This week, our team met with a board president and a board member as part of that early work.

For a nurse-founded company serving licensed professionals, this is a natural extension of our focus on making board matters easier to navigate. The public needs concerns to receive appropriate attention. Professionals need allegations to be evaluated on an accurate record. Board staff need workable processes for carrying out those responsibilities.

Our direction is to develop this work with the people who understand the daily demands of regulation. That means grounding system design in actual case workflows and evaluating whether proposed improvements make information easier to find, responsibilities clearer, and preparation for review more reliable.

The evidence supports investing in that work. It also sets a standard for it: systems should help boards account for each active matter, preserve the integrity of the record, and devote more attention to the decisions that require their expertise. BoardWise’s early work with boards and agencies is directed toward that goal.

Sources and research notes

This article draws on selected official reports and regulatory publications. The examples are not a representative national sample, and no national growth rate is inferred. Fiscal-year periods, intake categories, open inventories, and closed-case measures are kept distinct. California percentage changes were calculated from the reported endpoints and rounded to one decimal place. Audit snapshots and budget requests are identified by their reporting dates. The BoardWise update reflects the company’s account of its early work and meeting during the week of November 2, 2026.

1. [1] California Board of Registered Nursing, November 2025 board meeting materials, Table A and reference guide Open original source
2. [2] California Department of Consumer Affairs, FY 2026–27 Board of Registered Nursing Investigations budget change proposal Open original source
3. [3] Arizona Auditor General, State Board of Nursing performance audit and sunset review highlights, September 2025 Open original source
4. [4] Arizona Auditor General, Comments on the Board response, September 2025 (Appendix d, within the full audit report). Open original source
5. [5] Texas Board of Nursing, 2025 Annual Report, Operation Nightingale (Linked to the Board's publications page; the annual report file itself could not be retrieved during source checks.) Open original source
6. [6] New Mexico Board of Nursing, Compliance and Discipline Overview Open original source
7. [7] Federation of State Medical Boards, Workgroup on Regulation of Artificial Intelligence in Medical Practice, May 19, 2026 Open original source
8. [8] Virginia Office of the State Inspector General, DPOR Compliance and Investigations Performance Audit, September 2025 Open original source

Source check

• Checked November 4, 2026 against references [1], [2], [3], [4] and [8]: California intake figures (3,682 to 5,330; 7,739 to 8,317), the 64% referral and 193% unassigned-investigation increases, the 8.0 special investigator and $1.43 million request; Arizona's 56%, 2,177, 1,543, 51 high-risk, and 3,067 to 5,378 figures; Virginia's approximately 300 complaints awaiting review, 314 of 906 cases over 150 days, 107-day timeframe and 29 data-entry delays. The two California percentages recompute to 44.8% and 7.5%.
• Texas figures in reference [5] (367 open cases, 11 schools, 121 resolved) are reported as supplied; the annual report file could not be retrieved during source checks.