VBP

You are currently browsing articles tagged VBP.

Does WellPoint care about quality…or reducing cost?

May 19, 2011 in Health Insurance, Hospitals, Value-Based Purchasing | 2 comments

One large insurer is planning to begin paying hospitals based on quality.

“WellPoint is replacing the system it uses to help offset rising medical and other costs at hospitals in 14 states that serve its Blue Cross Blue Shield plans, which cover 34 million people. In recent years, it has raised its payments to those hospitals by an average 8% a year.

Under the new system, the company will pay increases only to hospitals that score high enough on a test based on 51 indicators of treatment quality. The indicators include whether the facility tries to prevent patients from relapsing after they leave the hospital, whether it follows a safety checklist and how satisfied the hospital’s patients say they are with their treatment.”

Does Well Point really care about quality?  The answer is maybe.

Improving quality of care could improve WellPoint’s bottom line.  If patients demand improved quality of care, implementing a hospital VBP system could attract more members. Further, WellPoint could just be altruistic and this may be an attempt to improve the health of its members (the Healthcare Economist is skeptical of this point).

It could also be the case the WellPoint does not care at all about quality.  High-quality hospitals will get the same annual increase they did before; low-quality hospitals will get less.  The chairman of the Federation of American Hospitals (FHA) accurately notes that hospital quality measures are far from perfect and are less-than comprehensive.  Nevertheless, even if the selected metrics measured quality inaccurately, certain hospitals would still receive lower payments and WellPoint would benefit either through increased profits or increased market share (by lowering premiums).

Rather than responding to pressure to increase qualityof care, WellPoint’s VBP efforts may in fact be a response to employer and beneficiary pressure to reduce premiums.

Tags: , , ,

Medicaid P4P

March 7, 2011 in Medicaid, P4P | 1 comment

As part of health reform, Medicare is looking to institute value-based purchasing or pay-for-performance programs in a number of settings.   In fact, in my work for Acumen, I have worked on a number of these initiatives (e.g., P4P for physician efficiency profiling, implementing a VBP system in home health).  Medicare, however, isn’t the only public insurance program to implement P4P.  Today, I provide an overview of State Medicaid P4P programs.  Here are some highlights from a report by Kuhmerker and Hartmann (2007).

  • As of July 1, 2006, more than half of all state Medicaid programs were operating one or more pay-for-performance programs. Within the next five years, if all current plans to start new programs are realized, nearly 85 percent of states will be operating Medicaid pay-for-performance programs.
  • Medicaid is not a new entrant to the field of pay-for-performance: almost half of all existing programs are more than five years old. A similar percentage of programs began operations within the past two years. More than 70 percent of planned new programs are expected to start within the next two years.
  • Seventy percent of existing Medicaid pay-for-performance programs operate in managed care or primary care case management (PCCM) environments, focusing on health care for children, adolescents, and women. While planned programs are still focused on managed care and PCCM providers, they appear to be shifting their emphasis to environments in which quality and cost issues related to chronic disease management can be better targeted. Rewarding the provision of primary care continues to be a component in the vast majority of Medicaid pay-for performance programs.
  • Nine Medicaid programs are joining with other payers, employers, consumers, and providers in statewide and regional pay-for-performance and quality improvement efforts.
  • Health information technology (HIT) is a focus of numerous Medicaid pay-for-performance programs. Several Medicaid programs are “paying for participation,” rather than “performance,” in an effort to encourage providers to adopt electronic health records, electronic prescribing, and other technologies.
  • The vast majority of Medicaid directors reported that their priority in operating pay-for-performance programs is to improve quality of care rather than reduce costs.
  • HEDIS and HEDIS-like measures are most popular in Medicaid P4P.
  • In 2000, 55.8 percent of all Medicaid beneficiaries were enrolled in managed care; by December, 2004, this percentage had increased to 61.3 percent. Managed care is the primary P4P setting for Medicaid. Primary care case management (PCCM) is the second most prevalent provider type included in P4P programs.
  • Almost all states use attainment or attainment and improvement scores to assess provider performance.

Of particular interest to an economist, P4P bonus payments are paid through a variety of different mechanisms. Examples include:

  • a maximum pool is established. If the provider performance payments would result in bonuses greater than that amount, the bonuses are prorated. If provider
    performance payments would not use the complete pool, only the amount calculated is distributed;
  • a pool is established and all providers meeting the necessary standard receive a proportional share based on their relative performance. The entire pool is distributed;
  • a bonus amount is established per occurrence. Bonuses are paid out based on the number of occurrences and the dollar amount per occurrence;
  • a bonus equal to a specific percentage of a reimbursement rate is paid when a standard is met;
  • the bonus is an established share of a calculated amount saved as a result of the P4P program (for example, in shared savings situations). The share is usually included in a contract between the state and the provider or vendor;
  • a bonus is calculated, but can only be used to offset any penalties; and
  • in recognition of CMS guidelines in this area, states often include provisions that ensure that no plan can receive more than 105 percent of their capitation rate as a result of any redistribution of, or increase in, funds.

Other incentive schemes include penalties, differential reimbursement rates based on past performance levels, increased probability of receiving an auto-assigned Medicaid beneficiary for good performers, withholds, and grants.  These payments are most frequently made in six-month or three-month intervals. To accommodate billing lags, validation activities, and other calculation-related processes, the time period between the conclusion of the measurement interval and when the incentive is actually received ranges from one quarter to one year. Non-financial incentives include tools, initial bid ranking, and public recognition.

To implement the P4P program, 90% of programs rely on information from providers.  In some states, programs contract with vendors to collect data additional data.  Fifty percent of state Medicaid directors reported that internal Medicaid staff conduct validation of program-related information themselves by sampling the data.  Thirty percent of respondents said that their state hires consultants specifically for data validation purposes.

Does Medicaid P4P work? Most State Medicaid Directors don’t know. Fifty five percent have not conducted formal, either because the P4P program was new or due to limited financial resources.

Source:

Tags: ,

Value-Based Insurance Design

February 11, 2011 in Health Insurance | 5 comments

There have been a lot of “value-based” initatives in health care of late.  Value-based purchasing aims to reward high-quality providers with either higher reimbursement rates or periodic bonuses.  On the other hand, value-based insurance targets the patient side.  The value-based insurance scheme varies cost sharing depending on whether the drug, procedure, or treatment is considered high value.  Disease management is also often linked with value-based insurance strategies.

For instance, this paper examined the effect of instituting a value-based insurance design for drug purchases.  The company’s health plan charged a “10 percent coinsurance for retail prescriptions and 7.5 percent coinsurance for mail-order prescriptions used to treat these three conditions. The coinsurance percentage was calculated as a percentage of the total cost of the medication. Medications for conditions other than those included in the program were assessed 20 percent coinsurance for retail and 10 percent coinsurance for mail-order prescriptions…General disease management programs for asthma, cardiac conditions, and diabetes were also implemented.”

By varying copayment rates, the company found that adherence rates increased significantly.

Another study found that disease management can greatly improve adherence.

V-BID Center

For those interested in Value-Based Insurance Design, I recommend you check out the following resource.

“The University of Michigan Center for Value-Based Insurance Design (V-BID Center) is the leading advocate for development, implementation and evaluation of innovative health benefit plans. Since 2005, the Center has been actively engaged in understanding the impact of value-based insurance design and collaborating with employers, health plans, policy leaders, and academics, to improve clinical outcomes and enhance economic efficiency of the US health care system.

Our website, www.vbidcenter.org, is designed to be a resource for employers, benefit designers, health care leaders, policy leaders and researchers interested in learning more about and implementing V-BID programs.  The site includes research (both by the Center’s faculty and other V-BID researchers), case studies, health reform and policy summaries, as well as presentations by Center faculty and upcoming events to learn more about V-BID.”

Sources:

Tags: ,

Physician Perspectives on Pay-for-Performance

January 26, 2011 in P4P, Quality | No comments

Survey Results

A 2004 poll of a sample of 400 randomly selected physicians found the following:

  • 71% of physicians supported payments based on the quality of care they provide,
  • over 60% of physicians thought that hospitals should also be paid based on the quality of care they provide,
  • Almost 90% of physicians thought that the current reimbursement system did not reward them for providing high-quality care, and
  • 62% supported public access to information about the quality of care they provide.

Quotation #1: Changing Perspective on P4P

“When I started my career in the late 1970s as an academic internist and geriatrician, I was skeptical of pay-for-performance, feeling that standards of care could not accurately assess the real benefits of my care of my frail elderly patients with multiple impairments. I still feel that way today because despite the rapid growth of the evidence base, we continue to measure relatively simple aspects of the process of care rather than measuring outcomes. I also felt then that pay-for-performance was a thinly veiled effort to increase efficiency rather than quality. Subsequent experience as the leader of a large academic health science center and as chief executive officer of a major health insurer has led me to believe that pay-for-performance holds substantial potential for enhancing quality of care.”

  • John W. Rowe, MD

    Quotation #2: The Problem of Complex Patients

    “Should metrics be simple or complex? Most current standards are simple. They state a basic clinical service that all patients with a certain condition should receive, such as prescription of β-blockers after myocardial infarction. Approximately 5% of patients are responsible for 50% of health care costs. They are typically complex. Therefore, we need standards to evaluate management of patients with chronic disease and multiple comorbid conditions. To achieve this goal, we will need a much richer evidence base than is currently available.”

    • John W. Rowe, MD

    Source: John W. Rowe, MD. Pay-for-Performance and Accountability: Related Themes in Improving Health Care. Ann Intern Med. 2006;145:695-699.

    Tags: ,

    Value-Based Purchasing in Nursing Homes

    December 13, 2010 in Medicaid, Medicare, Nursing Homes, P4P | 1 comment

    More than 3 million frail and disabled individuals rely on nursing home services in any given year.  About half of these individuals consider the nursing home to be their primary place of residence.  Nursing home quality, however, has often been called into question.

    Some experts believe pay-for-performance schemes will improve nursing home quality.  Today, I will review previous efforts to improve nursing home quality though P4P.

    Briesacher et al. (2009)

    A paper by Briesacher, Field, Baril and Gurwitz review P4P in nursing homes in a variety of states.   The authors note that “Approximately one-half of all Medicaid Programs currently operate some type of pay-for-performance program, and 85 percent have plans to do so within 5 years.”  This report consider nursing home P4P programs in: California, Colorado, Florida, Georgia, Illinois, Iowa, Kansas, Massachusetts, Minnesota, Ohio, Oklahoma, Texas and Utah.

    Measures considered include:

    • Clinical Measures (e.g., pressure sores, use of physical restraints, pain management, quality of life, MDS indicators, state-developed CAHPS measures, care plans)
    • Satisfaction Levels (e.g., patient, family, employees)
    • Structural Measures (e.g., nursing retention, staff turnover, occupancy rates, special licensure, state survey compliance, staffing hours/ratios)
    • Cost (e.g., Medicaid utilization, administrative costs, efficiency)
    • Pressure sores,
    • Use of physical restraints,
    • Pain

    Bonuses were paid depending on whether the nursing homes surpassed some threshold of these quality measures.  Some of the bonuses were paid as a flat rate ($3/day in Ohio and $0.50-$0.0 in Utah) and other used a percentage increase (up to 2.4% in Minnesota, 1%-3% increase in Iowa).  The highest bonus paid was 5% of per diem reimbursement where the lowest bonuses were $0.25.

    “We found little empirical evidence that pay-for-performance programs increase the quality of care of residents or the efficiency of that care in nursing homes. However, the program set in San Diego did find benefits, and it used the strongest of all evaluation designs, a randomized control design.” The San Diego RCT randomized nursing homes into treatment and control groups and gave the treatment groups incentive payments for: (i) accepting patients needing the most functional assistance, (ii) improving patient functional status, and (iii) prompt discharges of patients who remained out of the facility for at least 90 days.

    Read the rest of this entry »

    Tags: , ,

    Evaluating VBP Programs

    December 8, 2010 in P4P | 2 comments

    The call for the adoption of value based purchasing programs has gained in popularity in recent years.  These programs give physicians, hospitals, or other providers bonuses (or penalties) depending on the quality of care their patients receive. The Affordable Care Act (ACA a.k.a Health Reform) includes provisions to establish a VBP program for hospital payments based on hospital quality reporting; a national, voluntary, 5-year bundled payment pilot program; and a new payment structure for providers organized as accountable care organizations.

    Despite their popularity, evaluating the health outcomes or even documenting the processes required to produce positive health outcomes is difficult.  As these new VBP programs come online, it will be increasingly important to evaluate these demonstrations and identify best practices.  A paper by McHugh and Joshi (2010) makes some recommendation on how to improve evaluations of value-based purchasing programs.  A summary of their recommendations are below.

    Problem Recommendation Focus Target Audience
    Limited information on implementation and management of VBP programs Early and continuous collection of data on implementation Methods, infrastructure support Researchers, policymakers, providers
    Limited generalizability of findings More experimentation and greater variation in VBP Data, methods Researchers, policymakers, providers
    Lack of meaningful outcome measures Improved methods for risk adjustment, data validation, and measurement composition Data, methods Policymakers, providers, researchers
    Lack of integrated and aggregated data Support for EHR and HIT systems Infrastructure support Policymakers, providers
    Limited ability to synthesize learning from diverse VBP efforts Better practices and methods for synthesizing VBP program findings Methods, infrastructure support Researchers, policymakers, providers

    Most of these recommendations are sensible.  For instance, including data ‘checks’ to ensure valid collection of data would be useful.   Some suggestions, however, are more controversial.  In particular, the authors ask for “more experimentation and greater variation in VBP.”  Experimentation involves a tradeoff, however.  If I was a patient at a hospital under a VBP system, I would hope that the VBP program would be optimized given the current state of knowledge.  Experimentation could of course produce a superior system, but it could also create a worse one.  Thus, although increasing variation in VBP implementation would help researchers learn more and better understand if and under what circumstances VBP works, payers have a duty to make sure patient care is also optimized in the short-run.

    The authors give the example of Geisinger Health System’s ProvenCare which offers a single-episode price for CABG surgeries.  Geisinger’s integrated health system likely contributed to the success of this program.  In more decentralized health systems, should the goal be to implement a ProvenCare replica to see if it works in other settings or to design a VBP that is more tailored to the needs of the specific patients and providers it serves.  I’d tend to side with the latter strategy.

    Tags: ,

    Medicare Hospital Value-Based Purchasing

    December 7, 2010 in Hospitals, P4P, Quality | 4 comments

    The Health Reform (ACA) legislation mandated Medicare establish a hospital value-based purchasing (VBP) program by 2012.  In fact, the Deficit Reduction Act of 2005 already authorized Medicare to develop a plan to implement VBP for 2009.   How will they do this?  A CMS report from 2007 sheds some light on the topic.

    Since 2005, Medicare began the Reporting Hospital Quality Data for Annual Payment Update (with the incredibly unintelligible acronym of RHQDAPU).  RHQDAPU at first just required hospitals to report quality measures.  The Health Reform VBP initiatives, however, will begin to pay hospitals based on their performance on these metrics.  The 2007 CMS report claims that any VBP plan should contain the following 7 components.

    1. A Performance Assessment Model that is used to score a hospital’s performance on a specified set of measures, generating a Total Performance Score for each hospital.
    2. Translation of the VBP Total Performance Score into an incentive payment.
    3. A measure development process, including selection criteria for choosing performance measures for the financial incentive, and candidate measures for VBP Program start.
    4. A phased approach to transition from RHQDAPU to VBP.
    5. Redesigned data submission and validation infrastructure to support the VBP Program requirements.
    6. Enhancements to the Hospital Compare website to support expanded public reporting of performance results.
    7. An approach to monitoring VBP impacts, including potential impacts on health disparities.

    Below I discuss aspects of hospital VBP in more detail.

    Read the rest of this entry »

    Tags: , ,

    VBP Design Choices

    November 8, 2010 in P4P, Quality | 3 comments

    “Pay-for-performance (P4P) is one of the primary tools used to support healthcare delivery reform. Substantial heterogeneity exists in the development and implementation of P4P in health care and its effects.” Today, I review a paper which summarizes evidence, obtained from studies published between 1990-2009, concerning P4P effects.

    Measure Effectiveness

    Which types of measures produce the biggest change in physician behavior. The authors’ literature review reveals the following:
    “The effect of P4P on non-incentivized quality measures varied from none to positive.  However, one study reported a declining trend in improvement rate for non-incentivized measures of asthma and CHD after a performance plateau was reached.”

    In addition, process measures were more effective in changing physician behavior than outcome measures. Intermediate outcome measure effect of provider behavior was between the process and pure outcome measures. Among these measure types, programs where providers were involved in the VBP implementation lead to larger gains in outcomes. The authors do not mention if this was because of easier-to-game measure selection by providers or if this represented actual improvement. Providers can also game the system by declaring a patient ineligible for certain measures. “Gaming by over exception reporting and over classifying patients was kept minimal, although only three studies measured gaming specifically (e.g., 0.87% of patients exception reported wrongly). Therefore, there is limited evidence that gaming does occur with P4P use, although it is not clear what is the incidence of gaming without P4P use.”

    The most important factor may be whether the providers are aware that a P4P program has been put in place. “[S]tudies found positive P4P effects (5 to 20% effect size) with programs that fostered extensive and direct communication with involved providers.”

    Payment Size and Type

    The authors surprisingly found limited impact of P4P payment size on physician behavior. This may be due to the fact that in markets with payer fragmentation, even large P4P amounts will make up a small share of any one physician’s income. Generally, giving positive rewards produced better outcomes than programs with winners and losers, but the authors claim this finding is far from robust.

    Payments targeted to organizations seemed to less effective than those targeted at individual providers, but programs aimed at either tended to produce positive results. “A combination of incentives aimed at different target units was rarely used, but did lead to positive results.”

    Access and Equity

    Read the rest of this entry »

    Tags: , ,

    ESRD Quality Monitoring Initiative

    October 7, 2010 in Medicare, P4P, Quality, Uncategorized | No comments

    All patients with end stage renal disease (ESRD) are eligible for Medicare regardless of their age.  In 1972, the Social Security Act extended all Medicare Part A and Part B benefits to individuals with ESRD (of any age) who are entitled to receive Social Security benefits. ESRD beneficiaries now account for 1% of Medicare enrollment. This post review the types of services ESRD Medicare beneficiaries receive and how Medicare pays providers for these services.  Today, however, I will review Medicare efforts to improve the quality of care ESRD beneficiaries receive.  You can find an overview of the ESRD Quality Improvement Program (QIP) here.

    Timeline

    • 1972. the Social Security Act extends all Medicare Part A and Part B benefits to individuals with ESRD (of any age) who are entitled to receive Social Security benefits. ESRD beneficiaries account for 1% of Medicare enrollment.
    • 1978. ESRD Amendments require the formation of ESRD Network Organizations to support the ESRD program. CMS currently contracts with 18 ESRD networks.
    • 1994. The Core Indicators Project was established to improve the care of patients with ESRD.  The Core Indicators included measures related to anemia management, adequacy of hemodialysis, nutritional status and blood pressure control
    • 1999 (Mar). ESRD CIP was merged with the ESRD Clinical Performance Measures.
    • 2000 (Jan). Section 4558(b) of the Balanced Budget Act of 1997 required CMS to develop and implement a method to measure and report
      the quality of renal dialysis services furnished under the Medicare program. To implement this legislation, CMS developed the ESRD Clinical Performance Measures (CPM) Project based on the National Kidney Foundation’s Dialysis Outcome Quality Initiative (NKF–DOQI) Clinical Practice Guidelines.
    • 2001 (Jan). Medicare launched Dialysis Facility Compare based on the Nursing Home Compare website. The quality measures initially reported on DFC were measures of anemia control, adequacy of hemodialysis treatment and patient survival. Medicare claims data were used to calculate the anemia management and hemodialysis
      adequacy rates, and administrative data were used to determine patient survival rates.
    • 2008 (Apr). The updated ESRD Conditions for Coverage final rule, which contains revised requirements that dialysis providers and facilities must meet in order to be approved by Medicare and receive payment. As part of the revised requirements, dialysis providers and facilities are each required to implement their own quality assessment and performance improvement program. The CPMs were updated to include 26 measures from the areas of anemia management; hemodialysis adequacy; peritoneal dialysis adequacy; mineral metabolism; vascular access; patient education/perception of care/quality of life; and patient survival.
    • 2008 (Jul). Section 153(c) of the Medicare Improvements for Patients and Providers Act (MIPPA) requires that Medicare implement a quality incentive program (QIP)
    • 2009 (Feb). Medicare began implementing the CROWNWeb system to electronically collect information on about patients, facilities, providers, and clinical data to support the CPM Project.
    • 2009 (Sep). Medicare decides to begin paying ESRD providers based on a prospective payment system (PPS) beginning in 2011.

    Where are we now?  Medicare will begin paying dialysis providers through a PPS beginning in 2011.  This will give providers an incentive to provider services more cost effectively, but also potentially will give them an incentive to decrease the quality of care.  To ensure that ESRD beneficiaries receive the same quality of services under PPS as under a FFS, Medicare developed the QIP.  Below, I review the QIP in more detail.

    Quality Incentive Program

    The QIP requires Medicare to establish an ESRD quality program using the following steps:

    1. Select measures;
    2. Establish the performance standards that apply to the individual measures;
    3. Specify a performance period with respect to a year;
    4. Develop a methodology for assessing the total performance of each provider and facility based on the performance standards with respect to the measures for a performance period; and
    5. Apply an appropriate payment reduction to providers and facilities that do not meet or exceed the established total performance score.

    Medicare has already chose quality measures for the initial year.  Data from the following three measures will be submitted to CMS via ESRD claims.

    • Percentage of Medicare patients with an average Hemoglobin <10.0 g/dL (2%);
    • Percentage of Medicare patients with an average Hemoglobin >12.0 g/dL (26%); and
    • Percentage of Medicare patients with an average Urea Reduction Ratio (URR) >65 percent (96%).

    The numbers in parentheses represent the national performance rates for all dialysis providers and facilities based on 2008 data from the Dialysis Compare website.  Providers receive a score between 0-10 based on their performance on each measure.  Medicare has recently proposed a scoring method which subtracts 2 points for every 1 percentage point the provider falls below the initial performance standard (e.g., if the initial performance standard for a particular provider or facility for the Hemoglobin>12 g/dL is set as the 2008 national average rate (26%), then if that provider/facility had 28% of Medicare patients with hemoglobin levels>12 g/dL during 2010, the provider/facility would receive 6 points for its performance on the measure as 28% is 2 percentage points below the performance standard). The provider’s total score could be weighted evenly across all three scores.  Alternatively, some have proposed weighting the Hemoglobin <10.0 g/dL at 50% of the score and the other two quality measures at 25% of the provider’s score to put more weight on avoiding low hemoglobin levels.

    Payment will be based on the provider’s score.  THe proposed payment reduction scale is as follows:

    • 26-30 points: 0.0%
    • 21-25 points: -0.5%
    • 16-20 points: -1.0%
    • 11-15 points: -1.5%
    • 0-10 points: -2.0%

    In the future, Medicare will consider expanding the QIP program to include additional measures.  Quality measures considered include:  Kt/V, vascular access rates, bone and mineral metabolism, and access infection rates.

    Source:

    Tags: , , ,

    Value-based pricing

    November 27, 2007 in Academic Articles, International Health Care Systems, Pharmaceuticals | Permalink

    In February of 2007, the UK’s Office of Fair Trading (OFT) recommended reform to Britain’s current Pharmaceutical Price Regulation Scheme (PPRS). The PPRS sets maximum and minimum profit levels from the sale of branded drugs to the NHS. The PPRS allows companies freedom to set prices as they please on new substances, but restricts subsequent prince increases. A system of price cuts has also been instituted as well, yet it is likely that firms take these future price cuts into account when making their original pricing decisions.

    Reform

    Are there any other options? Simeon Thornton (Health Econ 2007) argues that a value-based pricing (VBP) scheme would be superior. In VBP, the NHS would pay pharmaceutical companies based on the value of the pharmaceutical to the patient base. One question is how value is determined. Thornton proposes cost effectiveness studies, which in effect means that the government or academics will determine the price.

    Pricing will also be allowed to vary by subgroup since people with certain diseases may benefit more from a disease than others. Also, incremental improvement will be encouraged since marginal improvements in treatment will receive higher payments.

    This program does seem to be an improvement. It is dynamically efficient since pharmaceutical companies will be paid more for more cost effective treatments. Further, if it turns out patients do not like the medicine and no one takes it, then NHS will not be paying a lot for failed drugs. Also, after generics are available, the price will adjust downward.

    In the static environment, the pharmaceutical company will capture all the consumer surplus since price will equal marginal benefit. However, as time passes and generics enter the market, a large consumer surplus will occur.

    The major impediment of this reform is the problem of any centralized system: information. How will the NHS determine cost-effectiveness? Will it be impartial? Will the conclusions be manipulable by interested parties? These questions are easy to answer theoretically, but very difficult to predict empirically.

    Tags: , , ,

    Newer entries »