Department of Labor Logo United States Department of Labor
Dot gov

The .gov means it's official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you're on a federal government site.


The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Consumer Price Index

Measuring Price Change in the CPI: Medical care

The medical care index is one of eight major groups in the Consumer Price Index (CPI) and is divided into two main components: medical care services and medical care commodities, each containing several item categories. Medical care services, the larger component in terms of weight in the CPI, is organized into three categories: professional services, hospital and related services, and health insurance. Medical care commodities, the other major component, includes medicinal drugs and medical equipment and supplies.

The following topics that are frequently confusing to the public are explained in this factsheet: expenditure methodology, health insurance, prescription drugs, professional services, and hospital services.

Expenditure methodology

General CPI methodology

The CPI measures inflation by tracking retail prices of a good or service of a constant quality and quantity over time. Tracking retail prices allows CPI to capture changes in out-of-pocket household spending over time. Each month, the various item indexes reflect the observed price changes, aggregating up to the all items CPI.

In the aggregation process, each item index is assigned a relative importance, or weight. The weight of each item in the CPI is determined using the Consumer Expenditure Survey (CE) which collects information from the nation's households and families on their buying habits (or expenditures), income, and household characteristics. Goods and services that consumers spend the most on will be the most heavily weighted. Additional information on CE and how weights are calculated and updated can be found in the CPI and the CE sections of the BLS Handbook of Methods.

Medical care methodology

The CE tracks consumer out-of-pocket spending on medical care, which is used to weight the medical care indexes. CE defines out-of-pocket medical spending as:

  • patient payments made directly to retail establishments for medical goods and services;
  • health insurance premiums paid for by the consumer, including Medicare Part B; and
  • health insurance premiums deducted from employee paychecks.

Employer paid portions of insurance premiums and fully tax-funded medical care (such as Medicare Part A and Medicaid) are not considered out-of-pocket, and therefore not used in weighting the indexes. 

While the weight of each CPI medical care related index is determined by out-of-pocket spending, price change reflected by the indexes measure the total reimbursement to medical care providers. This includes medical care payments made by private insurance companies, Medicare Part B, and Medicare Part D on behalf of consumers.

For example, in the physicians’ services index, we consider the price of an office visit to be the patient’s $20 copay, as well as the $80 insurance payment to the physician, for a total of $100. The $100 figure is used when calculating any price change.

The reason for this apparent discrepancy is due to challenges the CPI faces when pricing health insurance; this is described in more detail in the health insurance section. Medicaid and worker's compensation payer types are not eligible for the CPI since consumers do not make out-of-pocket payments to participate in these programs.

Note: Since the relative importance of medical care only includes out-of-pocket expenditures, its share in the CPI is smaller than its share of gross domestic product (GDP) and other national accounts measures. GDP includes reimbursements that are fully paid for by public sources and employers, increasing medical care’s share of GDP.

 Table A. Definitions of published medical care indexes and relative importance as of December 2018.
Item Definition Relative importance (percent) Percentage of the Medical Care Index

Medical care

Medical care commodities and medical care services 8.682 100%

A. Medical care commodities

Prescription drugs, nonprescription over-the-counter-drugs, and other medical equipment and supplies 1.707 20%

1. Medicinal drugs

All prescription and over-the-counter drugs 1.650 19%

a. Prescription drugs

All drugs dispensed by prescription. Mail order outlets are included. Prices reported represent transaction prices between the pharmacy, patient, and third party payer, if applicable. 1.308 15%

b. Nonprescription drugs

All nonprescription drugs, including topicals 0.342 4%

2. Medical equipment and supplies

Nonprescription medicines and dressings used externally, contraceptives, and supportive and convalescent medical equipment (e.g., adhesive strips, heating pads, athletic supporters, and wheelchairs) 0.057 1%

B. Medical care services

Professional medical services, hospital services, nursing home services, adult day care, and health insurance 6.974 80%

1. Professional services

Physicians, dentists, eye care providers, and other medical professionals 3.255 37%

a. Physicians' services

Services by medical physicians in private practice, including osteopaths, which are billed by the physician. Includes house, office, clinic, and hospital visits. (Excludes independent lab work and ophthalmologists. See Eyeglasses and eye care.) 1.732 20%

b. Dental services

Services performed by dentists, oral or maxillofacial surgeons, orthodontists, periodontists, or other dental specialists in group or individual practice. Treatment may be provided in the office or hospital. 0.785 9%

c. Eyeglasses and eye care

Services and goods provided by opticians, optometrists, and ophthalmologists. Includes eye exams, dispensing of eyeglasses and contact lenses, office visits, and surgical procedures in the office or hospital. 0.319 4%

d. Services by other medical professionals

Services performed by other professionals such as psychologists, chiropractors, physical therapists, podiatrists, social workers, and nurse practitioners in or out of the office. Also, includes independent lab work and imagining services. 0.419 5%

2. Hospital and related services

Services provided to inpatients and outpatients. Includes emergency room visits, nursing home care and adult day care. 2.621 30%

a. Hospital services

Services provided to patients during visits to hospitals, ambulatory surgical centers, or other similar settings. 2.340 27%

i. Inpatient hospital services (1)

Services for inpatients. Includes a mixture of itemized services, Diagnosis Related Group -based services, per diems, packages, or other bundled services. N/A

ii. Outpatient hospital services (1)

Services provided to patients classified as outpatients in hospitals, free standing services facilities, ambulatory surgery, and urgent care centers. N/A

b. Nursing home and adult day care services

Charges for residential care at nursing homes, nursing home units of retirement homes, and convalescent or rest homes. Also includes non-residential adult day care. 0.193 2%

c. Care of invalids, elderly and convalescents in the home

Fees paid to individuals or agencies for the personal care of invalids, elderly or convalescents in the home including food preparation, bathing, light house cleaning, and other services 0.088 1%

3. Health Insurance

Indirect approach based on retained earnings method. See Health Insurance  section. 1.099 13%

(1) Substratum index: a special index published below the typical item level. Relative importance is not available for these indexes.

Health insurance

Challenges to pricing health insurance

Even though insurance premiums are an important part of consumers’ medical spending, the CPI does not directly price health insurance policies. In a direct approach, we would track the movement of insurance premiums, holding constant the quality of insurance, and use these price relatives to build the Health Insurance index. However, the CPI has been unable to consistently control for changes in quality such as policy benefits and risk factors. Price change between health plans of varying quality cannot be compared, and any quality adjustment methods to facilitate price comparison would be difficult and subjective. As a result, we developed an indirect approach called the retained earnings method.

Retained Earnings Method

Methodology Framework

This section offers a general overview of the retained earnings methodology; the specifics of each step are detailed in later sections. This method begins by decomposing health insurance premiums into the two ways they are used by the insurance company: earnings retained by the insurance company and the benefits paid out on behalf of customers.

The earnings retained by the company can be thought of as leftover premiums income after paying out benefits. These earnings retained by the insurance company are used to cover administrative costs or kept as profit. The CPI thinks of this value as the cost of administering insurance services, such as paying out claims, and refers to it as the retained earnings.

Benefits paid out — or reimbursements to providers for medical goods and services — can be broken down as the product of medical care prices and medical care utilization. In other words, total benefits equals the average price of medical service claims multiplied by the number of claims filed.

Deconstructing health insurance premiums in this way shows that premiums are a function of retained earnings, utilization, and the price of medical care.

Figure A. Breakdown of Health Insurance Premium Components.

Figure A. Breakdown of Health Insurance Premium Components


Once premiums are defined, the CPI makes an important assumption: that the quality of a health insurance policy can be held constant through the ratio of retained earnings to benefits paid out. For example, if retained earnings rise as a result of increased premiums while benefits paid out stays constant, then the quality of insurance has decreased relative to its cost. This monthly relative of the retained earnings to benefits ratio is how the CPI tracks premium prices without having to adjust for quality changes in each insurance policy, such as services covered.

Since CPI only wants to track change in retained earnings in the health insurance index, CPI must reallocate the health insurance weight representing benefits paid out. This weight is reassigned accordingly to the non-insurance medical care indexes.

After this weight is reallocated, the monthly relative of the retained earnings to benefits ratio is multiplied by the monthly change in each non-insurance medical care index. This is done to account for the effect that changing medical prices has on premiums, and thus retained earnings.

We separate the process of reassigning weight and calculating the health insurance Index into four steps which are discussed in greater detail below.

  1. Separating health insurance index weights
  2. Calculating the retained earnings ratio
  3. Health insurance index aggregation
  4. Reassigning health insurance weight

1.      Separating health insurance index weights

The first step of the retained earnings method is to determine the appropriate weighting. The method we use reassigns part of the health insurance weight to the other medical care indexes. This is done by breaking down health insurance premiums, and thus the health insurance index weight, into the two ways premium payments are disbursed: insurance reimbursements for medical goods and services, and the administrative costs and profits of insurance companies (retained earnings).

Once separated, the weight corresponding to medical care reimbursements will be reallocated to the other non-insurance medical care indexes. The method for this reallocation is described in the reassigning health insurance weight section below. The weight corresponding to retained earnings remains in the health insurance index.

Figure B. Health Insurance Weight Redistribution.

Figure B. Health Insurance Weight Redistribution


2.      Calculating the monthly retained earnings ratio

The second step in the retained earnings method is to calculate the retained earnings ratio and its monthly relative. The retained earnings ratio is calculated using premium income and benefit payment data obtained from industry sources. Total premiums and benefits for each calendar year are sourced from the National Association of Insurance Commissioners (NAIC) and the California Department of Managed Health Care (DMHC) for commercial insurance. Non-profit premium and benefit data is obtained from national non-profit health insurance carriers. Prior to September 2018, commercial insurance data was sourced from A.M. Best’s Insurance rather than NAIC and DMHC. Since this information is not available in real time, the retained earnings ratio is generally lagged by 4 to 9 months, depending on when the industry data sources are updated.

The retained earnings ratio is calculated as:

  1. The total benefits paid out by the insurer is subtracted from its total premiums income to get the retained earnings.
  2. The retained earnings is then divided by total benefits to get a ratio of retained earnings to benefits.
  3. The annual relative change between ratios of retained earnings to benefits (the retained earnings ratio) is calculated by dividing the current year’s ratio by the previous year’s ratio.
  4. The annual retained earnings ratio is then converted to a monthly relative by taking its twelfth root. (See table B.)

Below is a hypothetical calculation of the change in retained earnings for commercial carriers:

Table B. Calculation for monthly retained earnings ratio.
Year Total Premiums Income Total Benefits Retained Earnings Retention-Benefit Ratio


$100,000 $94,000 $6,000 0.06383


$107,000 $100,000 $7,000 0.07
  1. Total benefits paid out in year T are subtracted from total premiums income in year T to get the retained earnings.


    1.	Total benefits paid out in year T are subtracted from total premiums income in year T to get the retained earnings.


  2. The retained earnings is divided by total benefits to get a ratio of retained earnings to benefits.


    2.	The retained earnings is divided by total benefits to get a ratio of retained earnings to benefits.


  3. The Retention-Benefit Ratio for year T is divided by the Retention-Benefit Ratio for year T-1 to get the retained earnings ratio.


    3.	The Retention-Benefit Ratio for year T is divided by the Retention-Benefit Ratio for year T-1 to get the retained earnings ratio.


    The 1.09666 relative change, or 9.67 percent, is the annual increase in the retained earnings ratio. In other words, assuming benefits paid out has remained constant, the health insurance company has retained 9.67 percent more premiums income than in the previous year.

  4. Spreading this annual change equally over 12 months is done by taking the twelfth root of the retained earnings ratio.


    4.	Spreading this annual change equally over 12 months is done by taking the twelfth root of the retained earnings ratio.


    This 1.007719 relative, or 0.77 percent, is the monthly retained earnings relative.

3.      Health insurance index aggregation

The third step in the retained earnings method is to multiply the monthly retained earnings relative by the index relatives of the other nine non-insurance medical indexes. This allows the CPI to account for the effect medical price change has on health insurance premiums. These nine indexes are: physicians' services, dental services, eye care, other professional services, hospital services, nursing homes, home health care, prescription drugs, and medical equipment. 

This results in nine health insurance index products, each factoring in medical care price changes. Each index product is weighted using the retained earnings weight left in the Health Insurance index. The overall Health Insurance index is calculated by aggregating these nine index products by their corresponding weights. 

4.      Reassigning health insurance index weight

The fourth and final step of the retained earnings method is to reassign the previously separated Health Insurance index weight to the other nine non-insurance indexes. Remember that the health insurance weight is split between medical care reimbursement and retained earnings. The weight representing medical care reimbursement is what gets reassigned to the other indexes. This weight is further split by health insurance type (commercial insurance and Medicare) and then redistributed to each medical non-insurance index in proportion with how much each health insurance type pays for that medical good or service.

For example, let’s say Medicare is responsible for one-third of the reimbursement weight in the health insurance index and half of Medicare spending is used to pay for physicians services. This would mean one-half of one-third of the health insurance index weight would be transferred to the physicians’ services index.

Figure C. Reassigning Health Insurance Weight.

Figure C. Reassigning Health Insurance Weight


This process occurs for each health insurance type and each of the non-insurance medical indexes until all of the weight representing reimbursements is redistributed. Industry data is used to determine the percentage each health insurance type pays for each medical item.

After the reallocation of expenditure on benefits, non-insurance medical care indexes carry the weight of both out-of-pocket payments and the insurance reimbursements for medical care. Therefore, each of these non-insurance indexes must track the price relatives of the total reimbursement (insurance reimbursements as well as the out-of-pocket patient copays). Continuing the example above, if the CPI is tracking the cost of an annual physical paid for by Medicare, we would track both the patient co-pay as well as how much Medicare pays.

Prescription Drugs

The prescription drugs index measures price change of drugs purchased with a prescription at a retail, mail order, or internet pharmacy. The tracked price is the total reimbursement to the retailer from the patient and all eligible payers for a single prescription. Payers are any entities who reimburse health care providers for the cost of medical services and/or goods. Eligible payer types for prescription drugs are: patient self-pay (cash), commercial or private insurance, and Medicare Part D.

Prescription drugs that are primarily consumed and paid for as part of hospital visits are included in the hospital services index. A deeper examination of BLS methodology for handling prescription drugs can be found here.

Item sampling

The prescription drugs index employs a streamlined sampling method; the pricing unit, or quote, is a specific prescription for a particular drug. When sampling drugs to price at a pharmacy, we obtain a list of the last 20 prescriptions dispensed and each prescription is assigned a percentage representing its likelihood to be sampled. The assigned percentage is equal to each prescription’s price divided by the total price of all 20 prescriptions. Using these probabilities, a prescription is selected; this process is known as sampling by probability proportional to size (PPS).

In this case, the “size” refers to the total reimbursement (patient payment and payer reimbursement) to the pharmacy for the prescription. The more expensive or more popular a drug is, the higher its sales numbers will be, and the more likely it is to be selected. This item selection process is the desired procedure and is attempted at every pharmacy selected for pricing. We work with respondents who cannot provide the necessary information to determine a workable sampling process, which may include selecting a drug via equal probability or other simplified methods.

Special patent loss procedures

When a brand-name drug in the sample loses its patent protection, consumers will often switch to the cheaper generic drugs that have entered the market. To accurately reflect the market, the CPI will resample all previous instances where the brand was selected because no generic existed at the time. We then resample between brand and generic versions of the originally sampled drug using corresponding probabilities proportional to the share of prescriptions sold at the pharmacy in the past 3 months.

To give enough time for the market to fully adjust to the new generics, the CPI waits approximately six months after patent expiration before implementing the resampling procedure. Resampling occurs only once. If a generic is selected, we treat any difference in price from the brand as a price change and the price difference is reflected in the index in the month when the resampling was performed. If the brand is selected again, we simply continue to price the brand.

Drugs changing to over-the-counter

If a specific prescription drug becomes available over-the-counter (OTC), the CPI will continue to price any quotes of the drug in the prescription drug index until it rotates out under normal rotation procedures (a complete rotation occurs once every 4 years with one-eighth rotating every 6 months). During this time, the quote remains in the prescription drug sample and any price change is reflected in the prescription drug index. Any quotes for that drug are not immediately transferred to the non-prescription drugs index, but once the drug becomes available OTC, it is eligible for selection in the non-prescription drugs index.

Recent changes

In 2016, the prescription drugs index changed its estimation formula from a geometric means formula to a Laspeyres formula. This was done following CPI research finding that the substitution effect for prescription drugs is overstated when using a geometric means formula.

Professional services

The professional services index covers services that are performed and billed by private-practice medical doctors, dentists, eye care providers, and other medical providers. Physicians' services and dental services indexes have most of the weight for this category. The professional services index collects prices on three different payer types: patient self-pay (cash), commercial or private insurance, and Medicare Part B.

Since Medicare Part B does not cover most dental services, the dental services index does not accept Medicare Part B payments, but does accept patient self-pay and private insurance.

Item sampling

The methodology for initiating all professional services is the same. The pricing unit is a doctor’s visit, defined by a specific medical service. At the initial visit, we establish the practitioner's specialty; if it is a group practice, one practitioner is sampled. Then a medical service is sampled via PPS. Here, “size” refers to the total reimbursement of each service over the previous year. For Physicians’ Services, Current Procedural Terminology (CPT) codes are collected to help describe the item accurately. Like most other CPI items, samples for professional services are rotated once every 4 years with one-eighth of the sample rotating every 6 months.

Recent changes

In September 2014, the CPI began pricing services at Health Maintenance Organizations (HMOs); previously HMOs were not eligible.

In June 2017, the CPI began using a carry forward process for some prices in the physicians’ services index in order to reduce the time burden on CPI survey respondents. Many physicians make limited adjustments to their prices each year, usually only changing prices when renegotiating their annual contract with insurance companies. This means physicians’ offices are eligible to have prices collected less frequently. For non-pricing months, the last collected price for each quote is carried forward for use in the current month index. Most offices that are approved for carry forward are priced between 2 and 4 months per year. The selected pricing months are those where price changes are most likely to occur.

In April 2018, the CPI began calculating sub-indexes at the payer type level for the Physicians’ Services Index. The CPI uses the Medical Expenditure Panel Survey (MEPS) to weight the associated payer type indexes within a geographical area. The CPI then aggregates these area payer-type sub-indexes to calculate the U.S. level change. This change was made to improve the index representativeness of the various payer types by increasing the share of payments covered by private insurance relative to payments made by uninsured people and payments covered by Medicare Part B.

Hospital services

The hospital services index tracks the price of services performed and billed by a hospital. It includes both inpatient and outpatient medical services, as well as room and board, lab work, and other services provided by the hospital. It includes services performed by physicians if they are employed by the hospital. However, it excludes services performed by physicians who do not work for the hospital, even if they are using hospital facilities. Those services are captured in the professional services index. Eligible payers for the hospital services index are: patient self-pay (cash), commercial or private insurance, and Medicare Part B.

Item sampling

The pricing unit is a hospital visit, defined by a specific medical service and a specific diagnosis or medical condition. At the initial visit, we work with the respondent to select a medical service using PPS. Here, “size” is referring to total reimbursement to the hospital. We then document the medical service and specific procedures of the hospital visit.

Reimbursements to hospitals can be categorized as either an itemized list or a lump sum payment. Reimbursement through an itemized list, usually called fee-for-service or fee schedule, results in a separate price for each individual good and service provided. Lump sum reimbursements, such as Diagnosis-Related Group (DRG) charges, result in a flat fee for all goods and services rendered based on the diagnosis or the type of procedure performed.

Recent changes

Following the August 2018 sample, the hospital services sample will be rotated less frequently. Outlets will be priced for 8 years instead of the standard 4 years. The sample will be static for 4 of the 8 years, after which the sample will undergo routine semi-annual rotation over the following 4 years. Currently the sample is in a static phase, the next wave of new outlets will be collected in February 2023.


Access data for medical care in our online database.

Additional information

We often receive questions about where potentially ambiguous items are categorized; table C contains a list of these items along with their expenditure category.

Table C. Ambiguous medical items and outlet types.
Item/Outlet Expenditure Category

Anti-Aging Cream

Personal Care Products

Appetite Suppressant

Medicinal Drugs

Asthma Inhaler

Medical Equipment and Supplies

Back Brace

Medical Equipment and Supplies

Breast Pumps

Miscellaneous Personal Goods

Chair for the Shower

Medical Equipment and Supplies

Compression Socks

Medical Equipment and Supplies

Dietary Supplements

Medicinal Drugs


Personal Care Products

Drug and/or Rehabilitation Centers

Hospital and Related Services


Personal Care Products

Eye Drops

Medicinal Drugs

Feminine Hygiene

Personal Care Products

Glucose Test Urine Strip

Medical Equipment and Supplies

Hair Loss Treatment

Personal Care Products

Independent Medical Laboratories

Professional Services

Intermediate Care Facilities for Individuals with Intellectual Disabilities

Hospital and Related Services

Latex Gloves

Medical Equipment and Supplies

Lip Balm

Personal Care Products


Personal Care Products

Massage Therapist

Professional Services


Personal Care Services

Medical Acne Treatment

Medicinal Drugs

Nasal Spray

Medicinal Drugs

Nasal Wash

Medical Equipment and Supplies

Needles for Injecting Medicine

Medical Equipment and Supplies

Nicotine Gum

Medicinal Drugs

Ovulation Kits

Medical Equipment and Supplies

Performance Supplement

Medicinal Drugs

Pregnancy Tests

Medical Equipment and Supplies


Medicinal Drugs

Protein Shakes

Other Foods

Shoe Orthotics

Personal Care Products

Snoring Treatment Strips

Medical Equipment and Supplies

Steam Inhaler

Medical Equipment and Supplies


Medicinal Drugs

Wander Alarm

Medical Equipment and Supplies

Experimental disease-based price indexes created by the BLS are available on the Price and Index Number Research page.

Additional information may be obtained from the Consumer Price Index Information Office by email or calling 202-691-7000. Information on the CPI's overall methodology can be found in the CPI section of the BLS Handbook of Methods.

Last Modified Date: April 24, 2019