Skip to main content
The BMJ logoLink to The BMJ
. 2002 Jun 1;324(7349):1332.

Getting more for their dollar: Kaiser v the NHS

Price adjustments falsify comparison

David U Himmelstein 1, Steffie Woolhandler 1
PMCID: PMC1123282  PMID: 12039832

Editor—The NHS is little cheaper than health care in the United States, according to Feachem et al.1 What's next on their agenda? War is peace? Freedom is slavery? The authors purport to show Kaiser's efficiency relative to the NHS. This task is hard, given two undisputed facts: firstly, the United Kingdom's per capita health expenditure is $1569, the United States's $4358; and secondly, Kaiser's casemix adjusted costs are about average for the United States. Undeterred, Feacham et al use an outrageous price adjustment, exclude many of Kaiser's costs, and ignore Kaiser's avoidance of the sickest and most expensive patients.

Feachem et al's price adjustment inflates NHS costs by 52%, assuming that the NHS plays no part in constraining drug prices, administrators' or specialists' incomes, etc. Conversely, the adjustment excuses the US system from responsibility for the world's highest drug costs and the billions wasted on healthcare executives and other hangers on. Feachem et al adjust away the price controls that are an important advantage of non-market systems.

Feachem et al trim Kaiser's costs by subtracting profits and high administrative expenses. Yet both are integral to the competitive market they extol. They falsely equate Kaiser's coverage with the NHS's, although Kaiser covers only miniscule amounts of nursing home care. Finally, many Kaiser patients—more than 12% according to a Kaiser memo—receive care outside Kaiser, costs which the authors exclude. Thus, Feachem et al understate Kaiser's actual costs.

Contrary to Feachem et al's assertions, Kaiser cares for a relatively inexpensive slice of the population. Their claim that Kaiser cannot avoid the expensively ill by booting them out is technically correct; when Kaiser members lose their jobs (for example, because of illness) and hence employer paid coverage, Kaiser must offer them individual policies. But Kaiser may charge whatever it likes—often thousands per month. Hence, few of the unemployed can actually afford coverage. Moreover, because the overwhelming majority join Kaiser through work, severely disabled people rarely get in. Although disabled Medicare patients may join, few do. Hence, Feachem et al's failure to adjust for casemix grossly biases cost comparisons; their adjustments for age and income are inadequate substitutes.

Finally, Kaiser's premiums (and costs) are virtually identical to those of other insurers that serve similarly healthy populations. Hence, Feachem et al's claim for Kaiser is tantamount to a claim that $1569=$4358.

The NHS has grave problems, and Kaiser is far from the worst of US health care. However, Feachem et al's conclusions are pure hogwash.

References

  • 1.Feachem RGA, Sekhri NK, White KL. Getting more for their dollar: a comparison of the NHS with California's Kaiser Permanente [with commentaries by J Dixon, DM Berwick, AC Enthoven] BMJ. 2002;324:135–143. doi: 10.1136/bmj.324.7330.135. . (19 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
BMJ. 2002 Jun 1;324(7349):1332.

Working knowledge would have been needed for comparison

David S David 1

Editor—I was very disappointed by the publication of the article by Feachem et al in comparing the California Kaiser system to the NHS.1-1 One has to live in California and be a doctor and have some working knowledge of the health maintenance organisation before passing any judgment about the quality of its care.

Firstly, Kaiser is not a “non-profit” health maintenance organisation. It is for profit, with the profit being divided between shareholder practising Kaiser physicians. The fact was made known by a Kaiser physician.1-2 So the profits shared are labelled as money spent on medical care for their patients, and therefore this justifies their classifying themselves as a non-profit health maintenance organisation.

Secondly, the data presented for their performance were collected and analysed by Kaiser personnel and may involve selection and information bias. I have recently responded to another outcome report by Southern California Kaiser regarding their observation that fewer studies and procedures results in better cardiovascular outcomes.1-3,1-4 My response and that of others exposed their self serving bias and distortions.1-5

Lastly, the commentary by Enthoven has to be interpreted with the knowledge that he was not only a consultant for Kaiser but also one of the directors of Blue Cross in California (another alleged non-profit organisation) and a leading proponent for the advent and proliferation of the healthcare delivery system that uses health maintenance organisations.1-1

I, like many of my colleagues in the United States, was hoping some day to have our country copy the much more humane and cost effective single payer universal healthcare coverage system in place in Europe and Canada for many decades. With the globalisation of the healthcare providers (pharmaceutical companies, insurance companies, health maintenance organisations, etc), acting through the auspices of the powerful World Trade Organization, Europe and Canada run the very real risk of acquiring our very inhumane, wasteful, and diseased healthcare system.

References

  • 1-1.Feachem RGA, Sekhri NK, White KL. Getting more for their dollar: a comparison of the NHS with California's Kaiser Permanente [with commentaries by J Dixon, DM Berwick, AC Enthoven] BMJ. 2002;324:135–143. doi: 10.1136/bmj.324.7330.135. . (19 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 1-2.York GK. Executives with white coats—managed care medical directors. N Engl J Med. 2000;342:130. [PubMed] [Google Scholar]
  • 1-3.David DS. Putting patients first. Cardiovasc Rev Rep. 2001;22:402. [Google Scholar]
  • 1-4.Mahrer PR. Outcome study of two large populations wih different rates of cardiac interventions. Cardiovasc Rev Rep. 2000;21:638–651. [Google Scholar]
  • 1-5.Weiss SR. Putting patients first. Cardiovasc Rev Rep. 2001;22:575. [Google Scholar]
BMJ. 2002 Jun 1;324(7349):1332.

Use of OECD database has led to incorrect conclusions

Nigel Edwards 1

Editor—The serious concerns that the economic analysis supporting the study by Feachem et al is fatally flawed has have been well explored by other respondents.2-1,2-2 There is a further problem, which is that Feachem et al have based their analysis of admissions and bed day use on the database of the Organisation for Economic Cooperation and Development (OECD). Anyone who is even slightly familiar with this dataset would know that it is important to triangulate it with other sources.

  • It is not clear that it always uses United Kingdom data—several variables are for England only

  • Data on admissions have been submitted in finished consultant episodes, not admissions

  • Data are incomplete

  • There are major definitional problems about acute care, day cases, long term care, etc. It cannot be assumed that these definitional differences, which have a major effect on numerous key variables in this analysis, have been controlled for by the OECD—Feachem et al make no mention of this.

The OECD database gives the same figure of 1 bed day per capita for each year 1993-8, which should have rung alarm bells. Triangulating these data with other sources further undermines the conclusion. In 1996 in England there were 120 232 beds, which gives a maximum number of bed days available at 100% occupancy of 43 884 680. This means that the maximum number of bed days per capita at 100% occupancy would be 0.897. In fact overall occupancy in the NHS in 1996 was much lower—there were not enough beds in England to give the figure Feachem et al quote. This is confirmed by hospital episode statistics that the bed days per capita were 0.644 for 1996-7.

Kaiser did better than the NHS on hospital use in 1996 and probably still does, but nowhere near as well as the article suggests. The factor that seems to have been ignored relates to the social care of older people—the NHS provides (somewhat against its will) a notable amount of social care in acute inpatient beds, and some adjustment should be made for this.

The lesson that Kaiser much more actively manages clinical processes and invests much more in making sure they work properly, has more specialists, and has invested in clinical leadership—which are the most important conclusions—could be lost by poor quality data analysis, dubious economics, and a tendency for all concerned to see in these data support for their own previously held convictions.

References

  • 2-1.Feachem RGA, Sekhri NK, White KL. Getting more for their dollar: a comparison of the NHS with California's Kaiser Permanente [with commentaries by J Dixon, DM Berwick, AC Enthoven] BMJ. 2002;324:135–143. doi: 10.1136/bmj.324.7330.135. . (19 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2-2. Electronic responses to Getting more for their dollar. bmj.com 2002. ( http://bmj.com/cgi/eletters/324/7330/135 (accessed 21 May 2002).
BMJ. 2002 Jun 1;324(7349):1332.

Kaiser may be model of American success or aberration

Jeremiah D Schuur 1

Editor—Feachem et al make a strong case for the efficiency and quality of the Kaiser Permanente system.3-1 Although comparing health systems is difficult, most health policy experts point to Kaiser as the best model of American health care. It is disingenuous to extend these results to other American managed care policies, as Feachem et al attempt to do. They say that managed care, of which the Kaiser system is one manifestation, is now the norm in the United States, covering 92% of all people with health insurance sponsored by an employer. Despite this, managed care has recently been criticised. Most members of health maintenance organisations, however, report satisfaction with their own health plans.

Their glowing report of Kaiser's success in a competitive market place raises several questions about the benefits of competition.

Firstly, although there is great financial incentive to decrease hospital use, there is almost no benefit for insurance companies to implement prevention programmes. The average US insurance plan has 25% annual turnover, discouraging such long term investments.

Secondly, Kaiser's quality has not resulted in market success. Their plan is not growing, and staff model health maintenance organisations are the exception rather than the rule in the United States. Instead of adopting plans similar to Kaiser most insurance companies are turning away from programmes with “defined benefits” to programmes with “defined contributions.”3-2 These programmes will require more out of pocket payments from the sickest people. There will be more financial barriers and disincentives to accessible care. For most Americans, there is no reason to believe that the future of health care in the United States will resemble Kaiser.

References

  • 3-1.Feachem RGA, Sekhri NK, White KL. Getting more for their dollar: a comparison of the NHS with California's Kaiser Permanente [with commentaries by J Dixon, DM Berwick, AC Enthoven] BMJ. 2002;324:135–143. doi: 10.1136/bmj.324.7330.135. . (19 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3-2. Freudenheim M. A new health plan may raise expenses for sickest workers. New York Times 2001 Dec 5; section a:1, col 6.
BMJ. 2002 Jun 1;324(7349):1332.

United States is paying more and getting less

John Robson 1

Editor—Feachem et al present disreputable comparisons.4-1 Unadjusted 1988-97 NHS figures for β blockers after myocardial infarction (42%), from one hospital in Britain's most deprived borough, are compared with 1999 Kaiser Permanente data extensively adjusted for contraindications.4-2 A more appropriate comparator (38%) was reported from Kaiser Northern California hospitals 1990-2.4-3 Similarly NHS retinopathy screening is inappropriately derived from a telephone survey of local estimates in 1996.4-4

Kaiser data exclude 15% of the population, recent entrants who were not members for a full year. In the table the NHS prevention data have been adjusted to exclude this group, which is assumed to have a rate 0.75 that of those full members. Adjusted NHS retinopathy screening for 1999 is 68% and NHS mammography and cervical screening increase to 71% and 91% respectively.4-5

Kaiser comparisons should include the 24% of Californians who are uninsured. The table assumes that the uninsured had 50% uptake for women's screening and 75% for childhood immunisation. Only 58% of Hispanic children in Los Angeles were immunised, and the completed immunisation rate for California as a whole was 75.3%. Not a single NHS health authority had rates as low.

The table does not show that Kaiser's preventive care or hospital productivity is better. NHS specialists do the same number of procedures per doctor as Kaiser, although NHS midwives ensure superior obstetric efficiency. The threefold difference in American heart surgery (or the twofold difference in France) is simply due to numbers of cardiologists. The article does not tell us that Kaiser doctors do not visit their patients at home, in or out of hours, or that Kaiser charges $10-15 for an office visit, $50 for casualty, and many other fees and copayments.

Table.

Comparison between the NHS and Kaiser Permanente*

NHS (England) Kaiser Permanente
Prevention (%):
 β Blockers after myocardial infarction 2 3 42 (42) 38 (93)
 Mammography 71 (69) 73 (78)
 Cervical screening 91 (84) 72 (80)
 Diabetic retinal examination5 68 (60) 68 (70)
 Three vaccinations with DTP and polio 95 (95) 88 (91)
Productivity (No of procedures):
 Angiograms per cardiologist 47.5 48.3
 CABG per cardiologist 58.7 52.9
 Caesarean sections per obstetrician 58 27
4-150

Numbers in parentheses are original data. NHS data exclude people within one year of changing practice; Kaiser data include uninsured people. DTP=diphtheria, tetanus, and pertussis. CABG=coronary artery bypass grafts. 

In 1997 annual per capita health spending was three times greater in the United States ($3724) and twice as high in France ($2125) than in the United Kingdom ($1193). Kaiser's facilities are more modern, better equipped, and have more support staff than the NHS. That costs more. There is much to learn about efficient use of hospital services, but underinvestment in the NHS remains the major problem.

America ranks 37th in performance of health systems, according to the World Health Organization; Britain ranks 18th, and France ranks 1st. The NHS serves all the people without payment at the time of use, without restriction, without exclusion or stigmatisation. Kaiser does not do any of that. When whole populations are assessed, a grossly underfunded NHS outperforms America by a wide margin.

References

  • 4-1.Feachem RGA, Sekhri NK, White KL. Getting more for their dollar: a comparison of the NHS with California's Kaiser Permanente [with commentaries by J Dixon, DM Berwick, AC Enthoven] BMJ. 2002;324:135–143. doi: 10.1136/bmj.324.7330.135. . (19 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4-2.Barakat K, Wilkinson P, Suliman A, Ranjadayalan K, Timmis A. Acute myocardial infarction in women: contribution of treatment variables to adverse outcomes. Am Heart J. 2000;140:740–746. doi: 10.1067/mhj.2000.110089. [DOI] [PubMed] [Google Scholar]
  • 4-3.Barron HV, Viskin S, Lundstrom RJ, Swain BE, Truman AF, Wong CC, et al. Beta-blocker dosages and mortality after myocardial infarction: data from a large health maintenance organisation. Arch Intern Med. 1998;158:449–453. doi: 10.1001/archinte.158.5.449. [DOI] [PubMed] [Google Scholar]
  • 4-4.Bagga P, Verma D, Walton C, Masson EA, Hepburn DA. Survey of diabetic retinopathy screening services in England and Wales. Diabetic Med. 1998;15:780–782. doi: 10.1002/(SICI)1096-9136(199809)15:9<780::AID-DIA632>3.0.CO;2-K. [DOI] [PubMed] [Google Scholar]
  • 4-5.Khunti K, Ganguli S, Baker R, Lowy A. Features of primary care associated with variations in process and outcome of people with diabetes. Br J Gen Pract. 2001;51:583–584. [PMC free article] [PubMed] [Google Scholar]
BMJ. 2002 Jun 1;324(7349):1332.

Like should be compared with like

Colin A McIlwain 1

Editor—As has been commented on, the business of making comparisons between health systems is difficult. In addition to the several issues already cited by respondents, the article by Feachem et al gives an NHS cost of £58.5bn but revenue allocations this year to health authorities in England are £37 157m.5-1,5-2

If the £58.5bn cited includes all NHS expenditure (and not just allocations to health authorities) in England then there are some very real comparative problems. Not included in the £37bn figure above, but possibly in the £58.5bn figure, would be central budgets of the Department of Health that fund—among other things—undergraduate nursing tuition and bursaries as well as those for allied health professionals and additional costs associated with teaching hospitals. These alone total nearly £1.5bn. Presumably the Kaiser figures do not include the costs of training and educating the health workforce in California. The £37 157m figure for health authority allocations includes among other things expenditure on public health and ambulance services.

Presumably Kaiser Permanente does not fund the public health departments at state, county, or municipal level in California or ambulance services in the state? The article also refers to the United Kingdom and uses its population. The Department of Health is the health department for England only and not for Scotland, Wales, or Northern Ireland. Nor is it the United Kingdom's department of health, as responsibility is split between the four constituent countries of the United Kingdom. The secretary of state for health is accountable to the British parliament but only for the money voted for use in England. It is not clear therefore whether the £58.5bn relates to England alone or is an aggregate of the funding in the four countries. Given the central conclusions in the article about the comparability between the NHS in the UK and Kaiser Permanente, it would be important for the NHS funding and population figures to clearly relate to one of the countries alone or the United Kingdom as a whole and for the NHS expenditure figure to be analogous to the services covered by Kaiser Permanente.

References

  • 5-1.Feachem RGA, Sekhri NK, White KL. Getting more for their dollar: a comparison of the NHS with California's Kaiser Permanente [with commentaries by J Dixon, DM Berwick, AC Enthoven] BMJ. 2002;324:135–143. doi: 10.1136/bmj.324.7330.135. . (19 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5-2. Electronic responses to Getting more for their dollar. bmj.com 2002. ( http://bmj.com/cgi/eletters/324/7330/135 (accessed 21 May 2002).
BMJ. 2002 Jun 1;324(7349):1332.

Length of stay is not the problem

Julian M Bene 1

Editor—Some readers of the article by Feachem et al have assumed that the NHS has three times Kaiser Permanente's hospital days per head because the average length of stay is that much higher.6-1,6-2 Explanations have occurred to readers, such as bed blocking because of poor social service care for recuperating patients, weak coordination of services in hospital, weak clinical practices, lack of emphasis on getting patients out quickly, etc.

Some validity to all of these inefficiencies is likely, but let's not jump ahead of the data. Feachem et al's table 3 shows length of stay of five days for NHS and four days for Kaiser. That's 20% more, not 200% more than Kaiser. Before assuming that the NHS must change to emulate Kaiser, let's recall recent American outrage at “drive by delivery” of babies and the legislation it provoked. Reduction of length of stay can be pushed to politically unacceptable extremes, even in the United States. Additionally, the push to limit length of stay can turn out to be an accounting misconception, if either costs of care migrate from inpatient to outpatient or that last day turns out not to cost much to provide, with minimal actual resource savings when it is eliminated.

The big difference between the two systems is not in length of stay but in admissions per head. Although the NHS may be admitting patients that Kaiser would treat on an outpatient basis, this cannot explain the huge gap. Almost certainly, Kaiser's population is much less sick than the NHS's. Indigent and trauma cases in the United States go to the county funded public hospitals, not to Kaiser. One suspects that Kaiser, acting in its own competitive interests, structures its benefit offering to elderly people to attract the healthiest and deter those who are likely to need hospitalisation. This may be true of employees, too, since employers offer them multiple health options with different premiums and copayments, and the Kaiser package may be designed not to appeal to people with known illness. Finally, given the intensity of health resources devoted to dying people, it would help to have a comparison of death rates in the Kaiser membership to compare with the United Kingdom's death rate.

If the health of the populations is as different as the admission rates suggest, then Feachem et al's conclusion may be invalid. Some of Kaiser's techniques may still be worth adopting in the NHS. But the despondency felt in Britain on hearing that the poor old NHS is not even efficient may be quite unwarranted.

References

  • 6-1.Feachem RGA, Sekhri NK, White KL. Getting more for their dollar: a comparison of the NHS with California's Kaiser Permanente [with commentaries by J Dixon, DM Berwick, AC Enthoven] BMJ. 2002;324:135–143. doi: 10.1136/bmj.324.7330.135. . (19 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6-2. Electronic responses to Getting more for their dollar. bmj.com 2002. ( http://bmj.com/cgi/eletters/324/7330/135 (accessed 21 May 2002).
BMJ. 2002 Jun 1;324(7349):1332.

Summary of responses

Alison Tonks 1

The paper comparing the NHS with California's Kaiser Permanente, a not for profit health maintenance organisation, concluded that Kaiser delivered substantially better care to its patients while spending no more per head than the NHS.7-1 It and its accompanying commentaries provoked an immediate and emotional response: we received 75 letters, 50 within a week of publication.7-2

Readers were outraged, puzzled, dismissive, disbelieving, even despairing. The outrage was levelled in equal measure against the investigators, for believing their own data, and the BMJ for publishing it. Respondents criticised the methods, the conclusions, and the solutions offered by commentators. They picked up flaws in the analysis, and some took a sledgehammer to the whole exercise. A few offered explanations for the apparent superiority of Kaiser; fewer still offered solutions for the NHS.

Respondents divided roughly into two camps: those who believed the data (27) and those who didn't (48). Comments came in from general practitioners, specialists, nurses, economists, health policy analysts, epidemiologists, government advisers, and even a risk analyst. Sadly, there were none at all from patients. Most worked in the United Kingdom, but a substantial minority were Americans. There were hardly any responses from Europe. Perhaps the BMJ's European readers don't see a fist fight between two failing ideologies as anything to do with them.

Several themes emerged from the paper's critics.

Firstly, adjustments made to the data were misleading, even wrong. The authors compared health care costs per head of population in the two systems and concluded that they were roughly the same. This result emerged only after several adjustments to the crude data.

Most controversial was an adjustment for the higher cost of drugs, staff, and services in the United States. In other words, the authors compared Kaiser's operating spend with how much the NHS would spend if it were operating in the United States. Which it isn't. So was the adjustment of “purchasing power parity” legitimate? At least 20 respondents thought not, referring to the adjustment as “sleight of hand,” “bizarre,” “outrageous,” “a serious flaw,” “a fudge factor,” and “a politically motivated abuse of statistics.” The same respondents also criticised the currency conversion rate, the fact that Kaiser's profits were excluded from the analysis, and the adjustments made for differences between the two populations. Hardly a single assumption remained unchallenged.

Secondly, respondents were dismayed by the authors' attempts to compare two such different systems serving two such different populations. For example, the NHS treats everyone, Kaiser treats only those that can pay its premiums (or can persuade the government to pay them). The NHS pays for health professionals' training and funds departments of public health, Kaiser does not. The authors made adjustments to account for these differences, and for differences between the Californian and British populations, but responders had no faith in them.

Thirdly, respondents questioned the real world value of the data on quality. The authors studied a range of quality indicators including waiting times, uptake of vaccinations, and cancer screening. But where was the patient's voice in all this? To paraphrase a handful of letters: Kaiser employs vastly more doctors than the NHS, and pays them nearly double what they are paid by the NHS. There's no evidence in this paper that Kaiser's service is proportionately better for patients. It's equally likely that the NHS is simply more efficient, delivering reasonable care at a lower cost.

In the end, 46 letters comprehensively dismantled the authors' analysis, and with it their conclusions. The details of the analysis can surely be defended (and have been) but the message implicit in many of these letters is that the authors and commentators let their ideology cloud their judgment. The same charge, of course, could be made against the paper's critics. The data are not robust enough to resolve the argument either way.

The 27 respondents who believed the data offered a variety of explanations for Kaiser's superiority. About a third mentioned that Kaiser Permanente had more of everything: more beds, more doctors, more nurses, more nurse practitioners, and better information technology than the NHS. About a sixth were also impressed by the integration of primary and secondary care. All these things, they thought, helped explain why Kaiser's patients in California spend so much less time in hospital than do patients in the NHS. Table 3 in the paper shows that Kaiser's patients occupy 270 bed days per 1000 population per year.7-1 The average for the NHS is almost four times higher (1000 bed days per 1000 population per year).

At least three letters complained that the NHS suffered badly from political interference, noting that 30 years of costly reorganisation has achieved little, if anything, for patients.

In general, the letters were more critical of Kaiser than the NHS. Many respondents simply did not believe that Kaiser could have so many more resources at its disposal, pay its specialists more than twice as much as NHS consultants, and still achieve similar per capita costs to the NHS. What about the “indigent poor,” they asked. What about the 24% of Californians without insurance? What about people with chronic mental illnesses? What about long term care? They concluded that Kaiser's patients must be richer, younger, and healthier than NHS patients.

The fiercest criticism came from a British nurse working for Kaiser in California. She complained of faceless insurance companies treating private “customers” as a resource and state funded “customers” as second class citizens. Others with first hand experience of insurance based health care included a medical social worker for Kaiser and a resident in emergency medicine from Providence, Rhode Island. Both wrote that Kaiser's results are not typical in a healthcare system that serves so many vulnerable groups so badly. One warned: “If your system serves the underclass, be proud of it.”

References

  • 7-1.Feachem RGA, Sekhri NK, White KL. Getting more for their dollar: a comparison of the NHS with California's Kaiser Permanente [with commentaries by J Dixon, DM Berwick, AC Enthoven] BMJ. 2002;324:135–143. doi: 10.1136/bmj.324.7330.135. . (19 January.) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7-2. Electronic responses to Getting more for their dollar. bmj.com 2002. ( http://bmj.com/cgi/eletters/324/7330/135 (accessed 21 May 2002).

Articles from BMJ : British Medical Journal are provided here courtesy of BMJ Publishing Group

RESOURCES