Psychotherapy, especially psychoanalysis and dynamic psychotherapy, has historically been poorly supported by insurance benefits. While there are a number of notable exceptions to this generalization (e.g., CHAMPUS, Medicare, Federal Employee Health Benefits Program in past decades), higher co-pays and lower yearly and lifetime limits for mental health care have been widespread discriminatory limitations. After years of advocacy for increasingly comprehensive mental health parity legislation, the Mental Health Parity and Addiction Equity Act (MHPAEA) enacted on Oct. 3, 2008, is the most sweeping national legal mandate to date for parity for mental health care benefits. The Affordable Care and Patient Protection Act of 2010 (ACA) also strengthened mental health parity and its official description explicitly lists psychotherapy as an “Essential Health Benefit.”
There are several reasons for the historically discriminatory pattern of under-reimbursement for psychotherapy. Stigma surrounding mental illness, as opposed to physical ailments, is well known and commonplace. There is also a widespread misconception about psychotherapy in particular, describing it as an unnecessary indulgence used by the affluent or the “worried well.” To operationalize bias against psychotherapy, insurers have historically resorted to numerous strategies such as coverage exclusions, unequal co-pays, and dubious “medical necessity” reviews. In justifying higher co-pays, for example, insurers have cited “price elasticity” meaning that a higher co-payment from an insured party is needed to reduce its use by patients to the same level as their visits to other medical care. The suppression of patients’ access to psychotherapy down to the same rate as use of other medical services by the deliberate increase in patients’ cost burden is conceptualized as a necessary correction for an assumed “moral hazard,” i.e., the unnecessary use of treatment by those not truly in need (Frank and McGuire, 2000).
Since MHPAEA made higher co-pays and discriminatory reimbursement schedules for mental health services illegal, insurance companies have resorted to more indirect means to limit their reimbursement. One such practice consists of flagging patients who exceed a pre-set, “outlier” session cap (e.g., more than 20 or more visits in a given period to trigger concurrent reviews that almost invariably deem continuing treatment “not medically necessary”). This offensive practice lacks both clinical support and an evidentiary basis. For example, with respect to the concern that psychotherapy is an unneeded, at best elective, expense, the Rand Health Experiment documented that even when psychotherapy is free, it is accessed by only 4 percent of an insured population and the average length of care is 11 sessions (Manning, Wells, Duan, et al, 1986). In fact, when insurance benefits for all care is generous, “price elasticity” and higher usage of outpatient mental health benefits tend to disappear. It also turns out that a higher cost burden for outpatient psychiatric care turns away very ill patients who simply forego treatment (Landerman, Burns, Swartz, et al, 1994; Simon, Grothaus, Durham, et al, 1996).
Exploiting a common prejudice that there is little evidence for the value of psychotherapy as part of a rationale not to cover its cost, insurers persistently overlook the robust research base documenting its effectiveness and disregarding in particular the studies of psychodynamic therapies. While psychological treatments are arguably more abstract and harder to conceptualize, to research and to measure than physical ones, we do have a compelling research literature documenting their efficacy and cost-effectiveness, including for psychodynamic treatments. Furthermore, the disparity in “medical necessity” determinations for mental health care compared to other medical care is also evident in the practice of providing insurance coverage for a substantial number of other medical services considered consistent with “standard practice”—and not necessarily supported by rigorous research or anything other than anecdotal or “expert opinion.”
Given decades of stigma and lack of appropriate support for psychotherapy and all mental health care, we now know this lack of sufficient treatment is a significant hidden multiplier of morbidity and disability, as well as greatly expanded overall health care expenses. Compared to patients without psychiatric illness, the increased medical expenses of the psychiatrically ill extend above and beyond the costs of their psychiatric care. Findings from a number of studies document these increased medical expenses, including more primary care visits, higher outpatient charges, and longer hospital stays (Melek and Norris, 2008; Luber, Hollenberg, Williams-Russo, et al, 2000; Deykin, Keane, Kaloupek, et al, 2001). Unfortunately, a high percentage of the psychiatrically ill are never even diagnosed and a majority of those who are receive inadequate treatment (Wang, Berglund, Olfson, et al, 2005; Wang, Lane, Olfson, et al, 2005). Simply put, patients with chronic, complex and/or recurrent psychiatric illness have more medical conditions and higher medical costs. These patients can often be treated with psychotherapy that yields better mental health and overall health outcomes. Yet these facts are unfortunately ignored by many insurance companies intent on minimizing reimbursement and evading the mandate for mental health parity. In addition, despite the preference for lower cost medication treatment, in many situations psychotherapy often provides a higher effect size than medication alone, augments the effect of medication (while the reverse is unclear), has lower dropout rates than medication alone protocols, and obviously lacks the side effects of medication treatments (Levy, Ehrenthal, Yeomans, et al, 2014).
Clearly any claim that a treatment should be covered under insurance must provide evidence for its effectiveness and, ideally, its cost-effectiveness. Cost-effectiveness is not synonymous with effectiveness or efficacy—it refers to the financial cost of a treatment and relates it to specific outcome measures of effectiveness (Cellini and Kee, 2010). It also does not mean cheap but what society is willing to pay for measurable positive outcomes. In essence, it signifies the impact per dollar spent. While many psychiatric patients improve with relatively brief courses of treatment, there are also important groups that are very costly to society if inadequately treated. Studies show that these patients often require more intensive and/or extended psychotherapy than most insurance companies are willing to support, despite the research that suggests the need for more care for these patients to achieve recovery as well as savings that often result from decreased medical expenses and improved productivity which is the even more stringent metric of cost-offset. However, insurance companies tend to focus on controlling short-term immediate costs and not on long-term planning and thorough treatment that might lead to better health outcomes and savings in the budgets of other parties.
Recent studies have delineated several diagnostic groups of patients who appear to need an intensive and longer duration of psychotherapy and who also often do better with a psychodynamic approach. These groups include those with chronic, debilitating personality disorders, those with chronic, complex disorders such as severe longstanding depression and anxiety, and those with multiple chronic psychiatric disorders. These patients are among the most seriously ill and are frequently not adequately treated with psychotherapy, due to arbitrary limits on reimbursement for psychotherapy by insurance companies (Bendat, 2014).
Patients with personality disorders have deeply ingrained, maladaptive, inflexible ways of thinking and behaving that generally lead to impaired relationships with others. Such patients are enormously costly to society. They are among the most chronically impaired groups in psychiatric populations: unemployed for longer periods, have more drug problems, suicide attempts, and interpersonal difficulties (Gabbard, 2000; Linehan and Heard, 1999; Pilkonis, Neighbors and Corbit, 1999; Reich, Yates and Nduaguba, 1989); more criminal behavior, divorce, child abuse, and heavy use of mental and general health care (Skodol, Gunderson, Shea et al, 2005). The lifetime prevalence of personality disorders is between 10 percent and 13.5 percent (P. Casey and Tyrer, 1986; Maier, Lichtermann, Klingler, Heun, and Hallmayer, 1992; Lenzenweger, 2008), affecting at least 30 million Americans of all social classes, races and ethnicities.
Randomized controlled trials of different psychotherapeutic approaches have demonstrated that dialectical behavior therapy (DBT), cognitive-behavioral therapy (CBT), psychodynamic and other specialized treatments for personality disorders are all effective, leading to reduced symptomatology, improved functioning and decreased hospitalization (Hadjipavlou and Ogrodniczuk, 2010, Bateman, 2012, Bateman and Fonagy, 2009; Linehan, Armstrong, Suarez, et al, 1991; McMain, Guimond, Streiner, Cardish, and Links, 2012).
For those who require an extended course of psychotherapy due to their mental illness, both longer duration and higher frequency of psychotherapy have independent positive effects. Together, these factors are associated with the most positive treatment outcomes (Rudolf, Manz, and Ori, 1994; Sandell, Blombert, Lazar, et al, 2000; Grande, Dilg, Jakobsen, et al, 2006; Leichsenring and Rabung, 2008, 2011). The factors that contribute to the cost-effectiveness of extended intensive psychotherapy for those patients who need it include savings from decreased sick leave and decreased medical and hospital costs (Dossmann, Kutter, Heinzel, and Wurmser, 1997; Bateman and Fonagy, 2003, 2008) (Bateman, 2003; Bateman and Fonagy, 2008); Clarkin, Foelsch, Levy, et al, 2001).
For those who require an extended course of psychotherapy due to their mental illness, both longer duration and higher frequency of psychotherapy have independent positive effects.
The disturbed interpersonal relationships of patients with personality disorders and other chronic complex disorders constitute a highly significant risk factor for increased mortality exceeding smoking, alcoholism, obesity and hypertension (Holt-Lunstad, Smith, and Layton, 2010). While psychotherapy of different approaches improves symptoms, a number of studies imply that long-term psychodynamic treatments are significantly superior in improving maladaptive interpersonal relationships (Huber, Zimmerman, Henrich, and Klug, 2012; Levy, Meehan, Kelly, et al, 2006; Levy, Ehrenthal, Yeomans, et al, 2014; Leichsenring and Rabung, 2008, 2011; Shedler, 2010). Shedler also showed that when compared to patients treated with other psychotherapies, patients treated with psychodynamic psychotherapy maintain therapeutic gains better and continue to improve after treatment ends, the “sleeper effect.”
With respect to patients with borderline personality disorder (BPD), Van den Bosch, Verheul, Schippers, and van den Brink (2002), found no empirical evidence that the core pathology of patients with BPD (unstable relationships, primitive defenses, identity disorder and boredom) is affected by one year of DBT. They also suggest that intrapsychic elements of this pathology may be more positively affected by psychodynamic psychotherapy. Levy, Meehan, Kelly, et al (2006) and Clarkin, Levy, Lenzenweger, et al (2007), also found that dynamic psychotherapy leads to broader personality changes than supportive psychotherapy or DBT for borderline personality disorder. And with respect to the need for longer term therapy, many studies indicate that patients with BPD in particular take significantly longer to improve (Howard, Kopta, Krause, and Orlinsky, 1986; Hoglend, 1993; Kopta, Howard, Lowry, and Beutler, 1994; Fonagy, 2002; Levy, Meehan, and Yeomans, 2010). In fact, the British Health Service National Institute for Health and Care Excellence (2009) cautions against the use of brief psychological interventions especially for borderline personality disorder stating, “…there is perhaps an even stronger signal that longer treatments with higher doses are of greater benefit. In several studies, significant improvement was only observed after 12 months of active treatment.”
While depression is the most common diagnosis made in primary care, Katon and Sullivan (1990) found that primary care physicians miss the diagnosis 50 percent of the time. It is experienced by one-fifth of all Americans at some point during their lifetimes (Kessler, Berglund, Demler, et al, 2003) and is extremely costly to society in increased medical costs, suicide-related mortality costs, and disability. A World Health Organization study (2008) found unipolar depressive disorders to be the greatest cause of worldwide disability. Compared to other depressed patients, the 20 percent who are treatment resistant have significantly greater health care costs, are twice as likely to be hospitalized both for depression and general medical admissions, have 12 percent more outpatient visits, 1.4 to 3 times more psychotropic medications, over six times the mean total medical costs, and 19 times greater total depression-related costs (Crown, Finkelstein, Berndt, et al, 2002).
For patients with unipolar depression, both a psychodynamic approach and the greater intensity of a psychoanalytic schedule appear to add benefit. A comparison of long-term cognitive-behavioral, psychoanalytic and psychodynamic therapy for these patients yielded similar improvements in depressive symptoms for all three approaches immediately post-treatment. While the CBT and psychodynamic therapy patients also had similar levels of depressive symptoms at three-year follow-up, patients treated with the more intensive psychoanalytic treatment had sustained greater
improvement. Both the psychoanalytically treated and psychodynamic therapy groups had fewer interpersonal problems than the CBT group at both post-treatment measurement points. The improvement in interpersonal problems was the only detectable superiority of psychodynamic therapy over CBT, while the more intensively psychoanalytically treated group had significantly greater improvement both in general distress and interpersonal problems immediately after treatment, and in depressive symptoms, general distress, interpersonal problems and self-schema than the CBT group at three-year follow-up (Huber, Zimmermann, Henrich, et al, 2012).
It seems increasingly clear that certain character traits complicate the treatment needs of depressed patients. For example, Blatt (1992) and Blatt, Quinlan, Pilkonis, et al (1995) found that perfectionistic patients do poorly in all brief treatments and fare better in more intensive, extended psychoanalytic treatment than in less intensive long-term therapies.
Also, seemingly successful treatment of depression is often accompanied by residual symptoms that can progress to become prodromal symptoms of recurrence. Residual symptoms have a strong prognostic value of relapse and are likely its most consistent predictors. Here again we find that dysfunctional social and interpersonal patterns in particular are positively correlated with persistent depression, relapse and poor long-term outcome. Accordingly, the fact that a patient no longer meets syndromal criteria is insufficient to designate full recovery despite the fact that the number and quality of sub-syndromal symptoms are often not specified in treatments judged to be successful. Therefore, treatments are needed that address ongoing characterological traits that put patients at risk for recurring illness (Fava, Ruini, and Belaise, 2007).
As noted above, a number of studies point to psychodynamic treatments as having greater efficacy with these traits (Levy, Meehan, Kelly, et al, 2006; Clarkin Levy, Lenzenweger, and Kernberg, et al, 2007; Leichsenrung and Rabung, 2008; Shedler, 2010; Huber, Zimmermann, Henrich, et al, 2012). Of note, demonstrating the impact of long-term psychodynamic psychotherapy on the brain, Buchheim, Viviani, Kessler, et al, (2012) published the first study documenting its treatment-specific changes in the limbic system and regulatory regions in the prefrontal cortex associated with improvement in depression after therapy.
Several researchers specifically link personality disorders with treatment resistant, persistent and recurrent depression and note that these patients with both major depressive disorder and personality disorder have significantly more role limitations due to emotional problems, impaired social functioning and general health perceptions than patients with major depressive disorder alone.
Poor psychosocial functioning compounds the impairments of major depressive disorder and affects the course of the illness. Furthermore, subjects whose personality disorders remit show improvement in social functioning and are more likely to achieve remittance of their depression than those with major depression and persisting personality disorders—the group that functions the poorest. These authors conclude that both personality and mood disorders need to be treated in comorbid patients, since the course of personality psychopathology influences depressive outcome as well as psychosocial functioning (Skodol, Grilo, Pagano, et al, 2005; Markowitz, Skodol, Petkova, et al, 2007). Depressed patients with comorbid personality disorders also have significantly longer time to remission than depressed patients without personality disorders. In fact, borderline and obsessive-compulsive personality disorders at baseline are robust predictors of accelerated relapse after remission from an episode of major depressive disorder, even when controlling for other negative prognostic predictors.
In sum, personality disorders are negative prognostic indicators for the course of major depressive disorder. Borderline personality disorder emerges as a particularly robust independent predictor of chronicity (accounting for approximately 57 percent of persistent cases) and also as the strongest predictor of persistence of major depressive disorder, followed by schizoid and schizotypal personality disorder, any anxiety disorder (the strongest Axis I predictor) and dysthymic disorder (Grilo, Stout, Markowitz, et al, 2010; Skodol, Grilo, Keyes, et al, 2011.) Taking the long view from a cost-effective perspective, it would seem clear that patients with major depression and a comorbid personality disorder need both illnesses treated to avoid recurrent and persistent depressive illness even when the treatment of the personality disorder may require a longer and more intensive treatment.
Other studies have examined outcome and cost-effectiveness for over 5000 outpatients with a variety of common DSM4 Axis 1 and 2 diagnoses who were treated with either long-term psychodynamic psychotherapy (LTPP) or psychoanalytic treatment. Both LTPP and psychoanalysis yield large effect sizes for symptom reduction, personality change, and improvement in moderate pathology both at termination and follow-up as well as reduced health care use and sick leave (DeMaat, Philipszoon, Schoevers, et al, 2007; DeMaat, de Jonghe, Schoevers, et al, 2009). Psychoanalysis, with its greater frequency, is more costly but more cost-effective than LTPP from a health-related quality perspective (Berghout, Zevalkink, and Haakaart-van Roijen, 2010a and b) and both treatments yield significantly reduced work absenteeism and lowered hospitalization at seven-year follow-up (Beutel, Rasting, Stuhr, et al, 2004.)
In addition, psychodynamic psychotherapies have been found to be effective for anxiety disorders, eating disorders, substance abuse, somatic symptoms and marital discord (Levy, Ehrenthal, Yeomans, et al, 2014).
In a succinct and accessible account of insurance coverage for mental health benefits in the context of MHPAEA and the ACA, Bendat (2014) describes both what these laws require as well as how they are circumvented and often fail to be enforced. Actual “parity” or equality for mental health benefits is mandated for insurance coverage for most medical insurance plans in both self-funded and fully-insured private employer plans if mental health benefits are offered and in both self-funded and fully-insured ACA plans with respect to essential mental health benefits with the exception of those “grandfathered” under the ACA. Recently, parity has also been expanded to mental health benefits in managed Medicaid and CHIP programs.
Parity regulations are meant to apply both to “quantitative” (number of services) and “non-quantitative” (describing protocols) limitations on the scope and duration of treatment eligible for coverage. Examples of non-quantitative treatment limitations include medical management standards, standards for provider admission to insurance networks and reimbursement rates, methods for determining usual, customary and reasonable charges, and “fail-first policies” that insist on lower-cost therapies prior to authorizing coverage for more expensive treatments. While an incomplete list, these standards and a number of others are prohibited from being applied in a more restrictive manner for mental health services than for other medical care.
Perhaps most frequently, however, the mandate for parity is being observed essentially in the breach. The evasions include the insurers rationing mental health care based on sub-standard and inappropriately restrictive medical necessity guidelines that are not developed by recognized mental health specialty groups, while adjudicating benefits for other medical conditions based on more generally recognized standards. To authorize more than a set minimum of mental health services, other illegal practices include, for example, a more restrictive insistence on fail-first treatment protocols and on much more severe and immediately life-threatening conditions (e.g., ongoing risk of imminent suicide) by which to evaluate requests for nonhospital levels of care. And in lieu of the older annual visit limitations and higher co-pays for mental health services commonly in use prior to MHPAEA, which the law has now proscribed, and in a hidden violation of the demand for parity in quantitative measures (number of services), insurers now use concealed algorithms to flag “outlier” patients who require more than a minimal, “normative” amount of treatment. These cases trigger the
ostensibly “non-quantitative” protocol of insurer reviews described as “quality control” to uncover “fraud and abuse” and limit care under the guise of “medical necessity” (Bendat, 2014).
To date, the processes meant to provide avenues for insured patients’ challenges to inappropriate denial of mental health benefits have been deeply flawed. Under Department of Labor rules, self-funded health plans (which cover nearly half of the country’s health benefits) are permitted to contract (generally through managed behavioral health care organizations) with “independent” review organizations (IROs) to adjudicate such consumer appeals with respect to benefit denials. IROs, however, routinely overlook parity and due process violations and very rarely reverse benefit denials since exercising actual independence and finding legal violations could compromise their contracts.
The CGRI is available to our members to aid with obtaining insurance reimbursement for our patients.
While the states have primary responsibility to enforce parity compliance of fully insured insurance plans, the states do not routinely scrutinize denials with respect to parity requirements and also routinely employ the same IROs who service the self-funded insurance companies, leading essentially to the same result. In practice, for-profit insurance companies put up a stiff resistance not only to covering the most expensive mental health benefits of hospitalization and residential treatment but also work vigorously to limit access to outpatient psychotherapy, particularly that which exceeds a brief course per year (Bendat, 2014).
Aside from these systemic obstacles inherent in the current system provided for appeals, in theory there always remains the potential remedy of litigation, however costly, financially and emotionally, for insurance consumers facing wrongful denial of coverage for needed mental health services. Individuals with employer-sponsored mental health benefits can exercise a private right to enforce parity and due process remedies conferred by MHPAEA. However, even though the parity requirements apply also to individual and non-federal governmental health plans regulated by the states, these subscribers lack a right to private legal action to enforce their entitlement to mental health care parity, thus limiting recourse to approximately 30 million insured subscribers (Bendat, 2014).
Among other measures, what is clearly needed are policy and regulatory revisions, the right of private legal action to all insurance subscribers, and establishment of true independence for “independent review organizations” adjudicating appeals of claim denials. What have been encouraging to date are large class action suits against offending insurance companies and managed care organizations successfully pressed by Meiram Bendat, among others. It is also important that national professional organizations such as APsaA join with other professional associations in policy statements and amicus briefs supporting genuine compliance with the requirements of the parity and affordable care act laws with respect to mental health care.
Both to protect our own patients and to arrive at a consensus within our own national organization to support parity officially, it behooves us as APsaA members to become familiar with the law and the many specific ways it is being evaded. In short, we need to get down a bit into the weeds about insurance practices such as inappropriate non-quantitative treatment limitations and spurious proprietary insurance company “medical necessity” guidelines not consistent with our own professional standards. According to Herbert Gross, chair of APsaA’s Committee on Government Relations and Insurance, the CGRI is available to our members to aid with specific problems encountered in obtaining appropriate insurance reimbursement for their patients. APsaA members may contact him at . In addition, on behalf of APsaA, the CGRI will work vigorously to support existing legal mandates for true mental health parity and to join in appropriate advocacy efforts to advance this cause on behalf of our patients.
We must insist on the recognition of our accumulated professional experience and wisdom, of the actual existent research base validating our crucial work, and of our patients’ suffering and financial burden from the lack of support for appropriate care.
Editor’s Note:
Much of the above material was also referenced in two volumes written by The Committee on Psychotherapy, Group for the Advancement of Psychiatry (GAP): Psychotherapy Is Worth It: A Comprehensive Review of Its Cost-Effectiveness, ed., Lazar, SG, APPI, 2010, a systematically searched, comprehensive review of 23 years (1984–2007)
of the medical literature relevant to the cost-effectiveness of all varieties of psychotherapy and a more recent survey of the research base, training, and practice issues in Psychodynamic Psychiatry Fall Special Issue: “Psychotherapy, The Affordable Care Act, and Mental Health Parity: Obstacles to Implementation,” Vol. 42, Number 3, Fall 2014, eds. Lazar, SG and Yeomans, FE. Details of the references cited can be found in these publications.
For further information on the references in this article, please contact the author at .