Sunday, March 31, 2013

A Primer on the Utilization Game

I want to post some references on the issue of "overutilization" but it is necessary to review the concept before I can post those references of make any further arguments about it.  Most people fail to understand that when they are talking about psychiatric practice in the US that it is tightly controlled by large health care and pharmaceutical middle men who make their profits to a large extent by denying care or insisting on cheaper care.  The very first articles using this term in medicine date back to the 1970s and involve policing various health care providers who were ordering unnecessary tests and procedures largely to prevent the loss of taxpayer dollars.  Some of the first articles looked at the problem as a combination of the need to assess quality of care according to certain standards, illegal behavior or intentional fraud, lack of education on the art of the practitioner, and "to ascertain where there is overutilization or underutilization of services perpetrated either by the practitioner or by the patient". 

In this early reference dental, optometry and podiatry services were an areas of focus and the measures of overutilization included too many x-rays, unnecessary fillings, unnecessary prescription of orthopedic shoes, and shorting prescriptions.  Professional services were evaluated by peer review and were categorized as being problematic because of unusual pattern of practice, poor quality of care,  unethical procedure, office facilities, qualifications for practice, abuse of billing codes, fraud, and self referral.  Although the source of the investigations and lack of equivalence of markers were problematic there ws a suggestion that overutilization was a significant problem.  Underutilization was suggested as a significant problem in under served populations but it was not systematically investigated.

The most systematic unbiased investigation of overutilization was done by the Peer Review Organizations in the late 1980s and early 1990s.  These efforts are documented to some extent in the National Academy of Sciences texts.  The protocol in the PROs consisted of a list of generic quality screens applied by nurse reviewers to hospital and clinic records.  The charts were also reviewed for appropriate utilization.  If a chart was flagged by a nurse reviewer it was sent to a physician reviewer for confirmation.  All physician reviewers were rigorously screened for qualifications and conflict of interest.  No reviewer could review records from any clinic or hospital that they were affiliated with.  Reviewers also had be in active practice and everyone knew that you could not make a living from reviewing charts for the PRO.

The result of the PRO experiment is a significant untold story.  A total of 6.3 million cases were reviewed using these protocols by 54 PROs across the country.  The denial rate for overutilization was 2.7%.  The frequency of quality problems was 1.3%.  The total cost of the program was about $300 million per year compared with the total cost of Medicare for the same year being $81.6 billion. I was a physician reviewer at the time and was eventually notified that the PRO program was being phased out because the cost of the program could not be justified by the amount of care denied ($300 million versus $220 million).  

What happens when overutilization is handled by companies that profit directly by denying care and the physician reviewers are either employees or contractors with that company?  As you might expect, the denial rate heads in a predictable direction.  Although it has not been extensively investigated, this article showed a denial rate of about 10% with rates varying with the companies involved.  As expected health plans with greater profit margins had higher denial rates and discounts.  Denial rates of 8-10% were replicated in another large study.  

At some point it became apparent to insurance companies that behavioral health services (their term for mental health and psychiatric services) would be an easy target for rationing and so-called "carveout" approaches.  This was buoyed by the Employee Retirement Income Security Act (ERISA).  ERISA effectively indemnified insurance companies and behavioral health plans against lawsuits over improper care.  Although there have been some suggestions that the courts may reconsider this indemnification, there has never been any significant movement in this area.  Managed care companies have successfully had their methods included in state statutes and have generally established a standard of care where rationing is a significant component.

A study by the Hay Group looked at the results of managed care rationing on mental health benefits as opposed to general medical benefits between 1988 and 1997.  There found a disproportionate decrease in mental health benefits across a number of parameters including:

- Fee for service plans were prevalent at the beginning of the study (92%) but they were largely replaced by managed care at the end of the study (20%)
- The value of general health care benefits decreased by 7.4% across the study but the value of behavioral health benefits decreased by 54.1%.
- As a total percentage of health care costs, behavioral health care decreased  from 6.1% in 1988 to 3.1% in 1997.
- Behavioral health care benefits were clearly rationed including a decreased number of inpatient days, a visit limit on outpatient care with per dollar visit limits and annual dollar limits that did not correct for inflation across the time of the study.
- Outpatient behavioral health care utilization decreased by 24.6%  between 1993 and 1996 while general health care utilization increased 27.4% in the same period.
- Inpatient mental health admissions decreased by 36.4% while general health admissions decreased by 12.7%.

The Hay Group Study was the best early evidence that mental health care was disproportionately rationed by managed care techniques.

If we fast forward to the present, managed care companies have taken the next step to make their rationing techniques as opaque as possible.  At some point some the largest companies have actually acquired the resources where health care is actually produced – clinics, hospitals, and groups of physician employees.   In that scenario they can bring their “overutilization” bias in house and use case managers to police doctors and tell them when to discharge patients.  The case managers are backed up by medical directors who are promoting the company line of a managed care company and who will do what they can to back up case managers if any physician is advocating for a longer length of stay.  They frequently have proprietary discharge guidelines that have not been scientifically validated that they use to establish discharge parameters.  It is no coincidence that the discharge dates all happen to be about the same time that most payers set as the maximum number of hospital days that they will pay for. 

The end result creates a health care system that is firmly entrenched to ration health care on the basis that there is an imaginary number of days or amount of money that can adequately treat a problem.  The only person who can advocate for the patient is their physician but he or she is clearly up against it.  The problem is more than being harassed by an outside company.  Now the physician’s job is on the line as well.  Disagreeing with the medical director on a consistent basis even a few times does not bode well for longevity within an organization.  In the case of hospital care we have physicians who realize that they need to discharge people in 4 or 5 days whether they have improved or not.  I can say from 22 years of inpatient experience that most people admitted to psychiatric hospitals with major psychiatric disorders do not improve to the point that they can be safely discharged in 4 or 5 days.  My conversations with outpatient physicians confirms this.  Typical managed care hospitals are no longer viewed as places where anything productive happens to improve patient stability.  The staff there will often admit it by saying that they are there for “mental health crises”.  But what happens when the crisis does not resolve in 4 or 5 days?

The limits on mental health care have also severely impacted outpatient care.  There is an emphasis on prescribing medication, often based on brief symptom checklists.  This also allows for the recruitment of large numbers of primary care physicians to treat problems once the checklist becomes the defacto mental health diagnosis.  Treating large numbers of people with anxiety and depression is much less expensive for health plans if the treatment is generic antidepressants or benzodiazepines.  Each patient is basically being “treated” for about $4/ month and they can be seen in follow up visits very infrequently.  It is well established in the research literature that different forms of psychotherapy work as well and in some cases better than medication for these conditions.  The research proven therapies generally require a specific course of treatment on the order of 8 – 20 sessions.  It is rare to see much therapy beyond three sessions in managed care settings and that would generally be received by a patient who was already taking a medication.

At this point we have devolved to a system of mental health care that devotes little time and effort to the treatment of mental disorders.  The treatment that does exist out there is clearly biased toward saving money for large health care companies who provide the bulk of it. All of that rationing is based on the premise that there is overutilization of services when the largest and best study shows that it does not approach the level of rationing that has occurred.

George Dawson, MD, DFAPA

1: Bellin LE, Kavaler F. Policing publicly funded health care for poor quality, overutilization, and fraud--the New York City Medicaid experience.  Am J Public Health Nations Health. 1970 May;60(5):811-20. PubMed PMID: 5462556; PubMed Central PMCID: PMC1348897
2: (1990) Medicare:A Strategy for Quality Assurance, Volume I: The National Academies Press.
3:  (1990) Medicare:A Strategy for Quality Assurance, Volume II: Sources and Methods: The National Academies Press.
4:  Hay Group: The Hay Group Study on Health Care Plan Design and Cost Trends, 1988 through 1997. National Association of Private Health Care Systems and National Alliance for the Mentally Ill, 1998.
5.  Dawson G.  The Utilization Review Hoax.  February 2012.

Tuesday, March 26, 2013

On the dangers of psychiatric diagnosis no longer being a process

I am inspired by a post on another blog having to do with the dangers of “premature psychiatric diagnosis”.  The author uses an anecdote to make a point about how a diagnosis of a psychotic disorder and then mismanagement of the treatment leads to a situation where there is no hope for the person affected.

From my perspective there are very few people with even severe psychiatric disorders who are hopeless.  In fact, people with some of the most severe cases of catatonia that I have treated became fully functional and were restored to their roles in their families and society.  That frequently occurred after months of inpatient treatment by a psychiatrist and staff who were interested and skilled in treating severe psychiatric disorders.  Much of what I did in 22 years of inpatient work was restoring hope and maintaining a hopeful atmosphere on my treatment team.

Diagnostic uncertainty is frequently cited as an area where mistakes are made.  Many studies document the medical comorbidity in patients with psychiatric disorders.  Despite anecdotal cases true medical causes of psychiatric disorders are rare.  I should qualify that by saying a brain disease, neurological or endocrine condition that is a direct cause for the psychiatric disorder is rare and I base that on screening patients and reviewing thousands of negative studies.  That said any acute care psychiatrist should know more about medicine and neurology than psychiatrists in outpatient settings because unlike their outpatient colleagues – they are responsible for making that determination.  On the psychiatric side, the potential list of causes of various syndromes is long and the actual diagnosis may not be evident until something happens on a long term basis.   A good example would be a drug induced psychosis.  In the ideal case, the patient is able to remain sober and any medical treatment for the associated syndrome can be tapered and discontinued.  In the real world, the chances of sobriety or even referral to a functional addiction treatment are low.
There are numerous limitations on psychiatrists.  The obvious one that practically all commentators leave out is managed care.  Is it reasonable to think that the diagnosis and treatment of any severe psychiatric problem like a psychotic disorder can be accomplished in 3 – 5 days?  That is the time frame that most managed care case managers are using to get people out of the hospital.  They often refer to purely proprietary guidelines on hospital lengths of stay that were clearly written by business people rather than clinicians.  I have been in the position of having a patient discharged by an administrator against my wishes so I know that it happens.  Managed care coercion is more subtle.  A managed care reviewer sitting at a desk in another state – reads chart notes and presumes to make a remote diagnosis and suggest that the person should leave the hospital.  They have no responsibility to the patient or their family.  Their only job is to get the patient out of the hospital to save the insurance company money.  Another constraint is at the level of public assistance.  Almost incredibly, many states link the availability of case management services to psychiatric diagnoses and they will clearly say in the statute that the person must have schizophrenia, major depression, bipolar disorder, schizoaffective disorder, or borderline personality disorder in order to qualify.  Having one of those diagnoses at discharge can be crucial to get housing and funded medical rather than be homeless.  That is a strong incentive to get the correct diagnosis sooner rather than later.

The work flow on inpatient units and in clinics is generally not considered.  If you have a psychiatrist seeing 12-15 inpatients and some outpatients and they are seeing 3 – 5 new patients a day that is not a lot of breathing room.  They will be (depending on other members of the team) able to collect collateral information from the family and outside sources, make direct behavioral observations, and relay treatment decisions and recommendations to the family.  In my experience occupational therapists, nurses, and social workers are all indispensable team members and often function in dual roles as a liaison with family members.  They can act as consultants to the family on legal and social issues as well as keeping them apprised of any changes in medical treatment on a day to day basis.

One of the key areas where care becomes fragmented both from a diagnostic and treatment standpoint is anytime there is a transition.  In terms of hospitals that occurs with any admission or discharge.  It also occurs between different outpatient clinics and between psychiatrists and primary care physicians.  I have been in situations where it took me two hours and calls to different physicians, pharmacies and relatives to reconcile a list of 10 medications.  At the end of that two hours I was still not absolutely certain of the patient’s correct medication list. 

The bottom line here is that good psychiatric diagnosis is a process. It is not like taking your car in to a mechanic and the mechanic plugging it in to an analyzer.  The best results occur when the patient and the family can communicate openly with the psychiatrist and any identified treatment team.  The diagnosis needs to take into account all of the available information and by definition it will only be as good as that information.   The critics of psychiatry always seem to think that this is a situation that is unique to psychiatric treatment.  As I have previously discussed it happens in all of medicine.  The basic difference being that many nonpsychiatric conditions lend themselves to analysis by a single observer.  There is something readily visible, audible or palpable that suggests an abnormality.  In psychiatry we are focused on communication, self report, and the observations of others.  We are also generally dealing with more information to make a diagnosis, especially if the patient’s capacity for self report is limited.  Psychiatrists more than anyone else need to be comfortable with diagnostic uncertainty and explaining these nuances to the patient and their family.

When the diagnosis is made it should be fully explainable to the patient and family.  Any stigma or negative reaction to the diagnosis should be discussed.  It should be evident that nobody is reducible to a psychiatric diagnosis given the fact that no two people are alike and each person is a unique individual with unique attributes.  This is true for any medical diagnosis and psychiatric diagnoses do not differ in that regard.   Nobody should leave the encounter with the idea that they are “hopeless”,  particularly in the case of a pure psychiatric diagnosis in the absence of a neurodegenerative disease.

I realize that most of us in one way or another are held hostage by a certain health plan, but if your psychiatrist or more probably your health plan does not follow that basic process – find a new one.  Getting stuck on whether or not a misdiagnosis has occurred without a plausible explanation for what has happened or continues to happen is generally not productive.  If you can’t get out of your health plan talk to the medical director and explain the deficiencies.  If that doesn’t work and you are concerned about the diagnostic and treatment process being rationed, contact your state insurance commissioner and file a complaint against the health plan.

Quality psychiatric care is possible, but it has been demonstrated that in many cases you have to fight for it.

George Dawson, MD. DFAPA

Sunday, March 17, 2013

More on Homicide Prevention – LA County Style

I have been developing a theme of how to prevent homicide and mass killing for more than a decade.  As previously posted, I think that this needs to be done independent of the firearms issue with a public health focus on both primary and secondary prevention.  There have been a couple of developments recently that I would like to highlight and whether or not they are consistent with the public health approach.

The first is an article in the NY Times today on a unique approach to school threat assessment and intervention.  The article describes LA County’s School Threat Assessment Response Team.  Several threat scenarios are described that trigger a multidisciplinary response from team members representing law enforcement, school officials, and mental health.  The way the program is described it is unique in terms of engagement.  Threats at school generally result in one dimensional and fragmented approach to the problem.  The school has a protocol that may result in suspension.  Referral to mental health providers is frequently a limiting step due to the lack of appointments, insurance problems, or debate over whether the school system or the health care system is responsible for assessment and treatment.  This patchwork system is a set up for people with severe problems falling through the cracks.

The LA County response is for the team to make a rapid same day assessment at the point of the threat and at the student’s home including looking at their room.  How many times have we read about the marginal teenager who is thrown out of school for threatening behavior and they end up sitting in their room focused on the same thought patterns or watching other forms of violent activity on the Internet or in video games?  Getting right into that environment seems like a powerful intervention to me and one that is likely to yield better results.  The main reason for failure in situations where a threat has been identified is that lack of follow up.    People who are threatening and aggressive are not likely to care if they are thrown out of school and they are not likely to follow through with mental health interventions.  The response team also spends time educating people about how to communicate in emergency situations where there are many misunderstandings about confidentiality.

The LA approach is innovative and exactly what is needed to assess and intervene in crisis situations involving threats and dangerous behavior.  In situation after situation, tragedies occur when people people come to the attention of someone and there is no clear map for assessment and treatment.  That is true in the school system, in colleges and universities, in the workplace and in family situations.  I have personally talked with people who said that they either did not know what to do or they actively tried several resources and were told that there were no appointments available or that the person was not dangerous enough to treat and unless they agreed to a voluntary assessment and treatment that nothing could be done.  But it doesn’t stop at that point.  I am also aware of situations where there clearly was enough evidence that the person was dangerous enough to meet criteria for an emergency assessment but it was not done of the person was released for the emergency department.  In many of these cases there was an adverse outcome.  What is the problem?

There is a significant bias against aggressive and violent people.  To some extent that bias is self protective.  Any reasonable adult knows the obvious advantages of avoiding conflicts or even irrational behavior.  There are always plenty of stories in the news about the lack of Good Samaritans in situations where an aggressive act is being perpetrated in public.  Many psychological explanations of this behavior are offered but I think the obvious motivation is avoiding the conflict and possible injury.  That same code of silence often applies in cases where there have been sudden changes in behavior and the person involved has a treatable problem.  A second level of bias is the moralistic approach to aggressive and violent behavior that equates this behavior with bad moral conduct.  That applies in situations where criminals use aggression to intimidate people and get what they want.  It does not apply when the aggression is a symptom of mental illness.

The bias extends beyond members of the general public.  The health care system is activated by a legal concept called “dangerousness” or “imminent dangerousness”.  Every state has different statutory requirements and those statutes are interpreted on a highly variable basis across every county in the state.  In some counties it comes down to some of the public officials involved seeing themselves as protectors of people’s rights.  In other counties, assessment and treatment are more of a priority.  At the level of the health care system there is another layer of bias.  The overwhelming bias these days is that people should not be assessed or treated in a psychiatric facility for more than 4 or 5 days and any assessment or treatment should be kept to the bare minimum.  It is easy to find different clinicians make entirely different decisions when presented with the same potentially dangerous patient.  The end result is a patchwork of acute care settings where people can go for help.  Because of all the biases involved unless an aggressive act has been committed the likelihood of an intervention occurring is basically a coin toss.

That is why the LA County response is so important.  It is an intervention that activates a rational response to threats from people who are likely in distress and possibly mentally ill.  There is no dangerousness standard initially and that is a critical departure from the current nonsystem.  The goal of the LA County response is to engage the person and their social network and not make a one-time assessment and decide to admit or discharge the patient based on a dangerousness concept.  The LA County response is unique in that it is based on behavior and the goal is to help the person involved rather than decide on whether or not they should be committed.  The overall approach is very similar to community psychiatry case management teams except LA County teams seem to have more latitude because they are not limited initially by commitment standards.

The is an excellent approach to the problem and I hope that it is researched, expanded to mental health crisis teams and widely adopted if effective.  I don’t know why it would not be effective.

George Dawson, MD, DFAPA

Erica Goode.  Focusing on Violence Before It Happens.  NY Times March 14, 2003.