Demonstrating performance to your board

"Boards exist to ensure that an organization is accountable to meeting the needs of the community or cause that the organization serves," said Denny Morrison, PhD, a former behavioral health CEO who hosted the session on behalf of Netsmart, where he now serves as Chief Clinical Officer. Boards function as an organization's "owner," with total accountability and authority until they choose to delegate or give it away to one or more employees.

 

Often, a behavioral health CEO is the board's only employee. The CEO is the top executive, with powers that, though broad, may be exercised only within limitations set by the board.  Citing a governance model by John Carver, Morrison explained that a critical board responsibility is to set a boundary between its role and that of the CEO. The board must set expectations for the CEO, then track the CEO's accountability to those expectations. Other than the CEO, a board does not interact with other organization managers with two possible exceptions. Boards may interact with a compliance officer and HR officer on matters specific to the CEO's performance.

 

Setting and maintaining a clear boundary between expectation and execution is a critically important board responsibility, Morrison emphasized. Execution is a management responsibility, ultimately accountable to the CEO. This means that the board must focus on the what of its expectations - explaining what the CEO must do to achieve success. It must not interfere or drift into how a CEO and management team achieves it.

 

Quoting Carver, Morrison stated, "A board should do all that it must do, not all that it can do."  Board members must avoid doing what they so often do in their non-board careers: they must not become managers.

 

 

Then, Morrison settled into the details of what types of performance management measures would enable the board to fulfill its governance role, specifically to ensure proper CEO and organizational performance. He defined performance as "a reasonable interpretation of the board's objectives."

As gatekeepers of all organization performance data, the CEO and management team must strike a careful balance in providing data to the board. Too much causes overload or misunderstanding, while too little may be perceived as "trying to paint a rosy picture," Morrison continued. 

 

Inevitably, he said, providing meaningful data to board members "involves getting the board tuned into relevance of certain organizational and business information and then, getting them to say so." The process of definition can be a challenging, but being as clear as possible in both the choice of data and its presentation is essential. 

 

Good board data are: 

Regular in frequency

Clearly explained and free of jargon

Relevant and balanced

Presented graphically when possible, with appropriate explanations

Open to detailed questions and answers

 

Good data, he added, generally incorporate two key features. "You all know the bell curve, with the majority of results clustered in the center, then fewer at each extreme? The key with boards is to define the most relevant types of data and then, within each, define what the "center" is - the usual results or data. Effective board discussions often present and review typical or usual results, but focus special attention on "the outliers," the results that aren't within acceptable limits, said Morrison. Often, boards will choose to review such outliers for several meetings until they appear to be addressed. Many types of financial data are reviewed in this manner:

 

Current ratio

Net op margin

cash on hand

Days in A/r

Bad debt

MGA as a percent of total expense

Unit costs  (Key consideration:  how to allocate "overhead" or indirect costs)

Revenue mix

 

"Bad," or ineffective board data is often typified by too much verbal explanation and too little numerical or graphical information. "Sure, your explanation is going to go in the meeting minutes, but you've got to offer up meaningful data."  At times, it may seem that some boards or members can't identify what data they really need, so it is vital to persist in presenting meaningful information and developing a shared understanding of it with board members. "Don't wait until it's too late to find out what your board really needs."

 

Morrison recommended that for clinical data, first divide it according to the level of care that is involved:  emergency care, essential care (required to prevent decompensation), and elective care. Then, he suggested focusing on essential outcomes only for board review, for example, if a program's objective is to reduce hospitalizations, than that should be the focus of the board's review.

 

Satisfaction measures are essential to understanding patients' reasoning for a premature exit from care, Morrison continued. And, though many patients lack the experience to judge the quality of behavioral healthcare, they will judge care quality based on their customer experiences in other contexts. "They'll use surrogates for quality," he suggested.

 

"Service satisfaction is usually based on the package in which its delivered," Morrison said.  For example, no show rates teach that patients are often shopping for what they want to buy on their first visit: can they get to the site of treatment? Do they like the people and the surroundings?  Will appointments fit in with personal, family, or school schedules?

 

Operational measures include: 

Access 

Productivity

Utilization--empty beds or bed ratios.

Client/Staff ratios

Measure of risk

Discharges

Hospitalization status

No show rates

 

With regard to measures of employees or organizational satisfaction, generally involving employee surveys, his advice is simple: "Don't ask if you're not going to reply."

 

Finally, Morrison indicated that when it comes to performance measures, they're valid only in the context of past performance.  "Data standing alone, without context, is meaningless," he said.

 

To gain the context required to make performance data relevant and meaningful, Morrison suggested benchmarking, which he sees as "the key to context and control" of organizational performance data. Practiced in industry for decades, benchmarking involves the capture, aggregation, and reporting of data from one organization that is then compared to data from other related or comparable organizations. 

 

But, organizations often face a significant barrier to benchmarking:  "The biggest problem in healthcare is just getting the data into the system," he asserts. Too often, would-be benchmarkers are unable to meet this basic assumption. 

 

For those wishing to benchmark, Morrison suggests:

 

manual data collection is acceptable, but computerized is faster.

the larger the data set collected (in number) the finer the measurements that can be made. 

3) sharing learning opportunities is essential to any benchmarking effort

4) learn best practices from top performers through private discussions.

 

Benchmarking has only one drawback:  While it tells you how you're doing relative to those benchmarked, the process doesn't do anything to explain the "why" of success or failure.  To do this, a more sophisticated measurement is required.