Clinical practice patterns of generalists and specialists in Alzheimer's disease: What are the differences, and what difference do they make?

Abstract

Optimising the roles played by both generalists and specialists in the diagnosis and treatment of Alzheimer's disease (AD) could have a major impact on the quality and cost of patient care. Therefore, one aim of the IMPACT survey was to characterise the similarities and differences between these 2 categories of physicians, in 5 different European countries, across a number of domains relevant to the medical care of people at risk for AD and those with the disease. Physician respondents comprised 250 generalists and 250 specialists from 5 European countries—France, Germany, Italy, Spain, and the United Kingdom. A substantial majority of generalists were either general practitioners or family physicians; the majority of specialists were neurologists. In April and May 2009, physician respondents completed a 30-minute, Web-based questionnaire during which they were presented with a number of multiple-choice-type questions concerning their knowledge of AD, approach to diagnosis and treatment of AD and experience of providing care for people with dementia. Generalists reported that 45% of their AD patients had mild symptoms at the initial visit compared with 60% for specialists (P<0.001). Specialists claimed that they diagnose patients with AD themselves in 65% of cases versus 33% for generalists (P<0.001). The main prescription treatment options employed were AD-specific medication (90%) and medication for mood or behaviour (78%). A similar percentage of generalists and specialists (77% and 75%) initiate drug treatment within 1 month of diagnosis. Overall, there were more similarities than differences between specialists and generalists regarding a broad spectrum of issues relating to AD; differences between countries appear to be greater than differences between physician groups.

Introduction

Optimising the roles played by both primary care physicians (generalists) and secondary care physicians (specialists) in the diagnosis and treatment of Alzheimer’s disease (AD) could have a major impact on the quality and cost of patient care (1). Currently, there is no clear “best practice” in this regard, although recent reviews have attempted to define the roles of generalists in dementia care (2, 3). What prevails is a widely varied pattern of health care service utilisation by patients with cognitive complaints and those diagnosed with AD (4). Different types of physicians, with various training backgrounds, working in a range of clinical settings, provide diagnostic and treatment services in varying degrees of coordination with each other and with non-medical service providers. In particular, considerable controversy exists regarding the specific roles that should be played by generalists and specialists in the diagnosis and management of AD, and how such medical care could best be organised and integrated with public health, mental health and psychosocial service providers (2, 5).

Therefore, an aim of the IMPACT survey was to better characterise the similarities and differences between these 2 broadly defined categories of physicians, in 5 different countries, across a number of domains relevant to the medical care of people at risk for AD and those with the disease. A separate article in this issue by Martinez-Lage et al discusses the attitudes and perceptions surrounding AD amongst physicians, as determined in the IMPACT survey; this article focuses on the IMPACT results that address physicians’ knowledge about AD and their description of the care they provide for patients with AD in practice. Key topics covered include knowledge of disease symptoms, knowledge of community resources, prescribing practices and referral patterns. These data are intended to help develop a clearer picture of the present state of the roles of primary care and specialist physicians in dementia care in 5 European countries, and include country-specific differences. The findings have potential implications for public health policy, health care delivery organisation and the training and continuing education of physicians.

Methods

Physician respondents to the IMPACT survey comprised 250 generalists and 250 specialists from 5 European countries—France, Germany, Italy, Spain and the United Kingdom (Table 1). For the purposes of this survey, the term specialist refers to those physicians whose field of specialization includes AD; the term generalist refers to physicians whose field of specialization encompasses general adult medicine, or who do not have specialty training. Generalists included general practitioners, hospital general physicians, family physicians and internists; specialists included neurologists, geriatricians, neuro-psychiatrists, psychiatrists and psychogeriatricians. A substantial majority of generalists were either general practitioners or family physicians; only 28 of 250 were hospital general physicians or internists. Similarly, a substantial majority of specialists were neurologists, except for in the United Kingdom, where there were only 20 neurologist respondents, with the remainder of specialists divided amongst the 4 other specialist categories. Physicians enrolled in the survey were required to be aged between 25 and 69 years, to have been in practice for between 5 and 30 years (3 and 30 years in Spain) and to currently spend more than 50% of their time in direct patient care. No physician respondents were employed either full time or part time by a pharmaceutical company. Enrollment was restricted to generalists who saw at least 2 patients with AD each month. Likewise, specialists were required to see at least 10 patients with AD each month (8 in Spain).

Table 1.

Physician Specialty

Respondent type	Total	France	Germany	Italy	Spain	UK
Generalists (GEN)	250	50	50	50	50	50
General practitioner/	222	37	42	49	46	48
family medicine
Internist	13	3	7	1	1	1
Hospital general physician	15	10	1	0	3	1
Specialists (SP)	250	50	50	50	50	50
Neurologist	195	45	37	46	47	20
Geriatrician	19	4	1	2	1	11
Neuro-psychiatrist	14	1	10	1	1	1
Psychiatrist	11	–	2	1	1	7
Psychogeriatrician	11	–	–	–	–	11
Total	500	100	100	100	100	100

Throughout April and May 2009, physician respondents completed a 30-minute, Web-based questionnaire during which they were presented with a number of multiple-choice–type questions concerning their knowledge of AD, their clinical practice, their approach to diagnosis and treatment of AD and their experience of providing care for people with dementia. Although every item was presented in a strictly multiple-choice response format, some questions did have more than a dozen responses from which to choose.

Responses were analysed for the total physician group, by country and by physician group (generalists vs specialists). The outcome measure was the percentage of each group endorsing a particular response.

Statistical significance testing was performed for selected items using a standard software program for the interpretation of survey data (Quantum, http://www.spss.com/software/data-collection/quantime/quantum.htm). Only significant P-values are shown.

Survey items pertaining to knowledge of AD and clinical practice

Knowledge

Respondents were asked to identify each of 20 response items as belonging to 1 of the following 4 categories: early signs of AD, later signs of AD, normal ageing or none of these. Respondents were not limited in the number of items they could identify as belonging to each category (Tables 2A, 2B).

Table 2A.

Signs of Late AD

	Frequency item chosen (%)
Response items (n=20)*	Total			France			Germany			Italy			Spain			UK
	All	GEN	SP	All	GEN	SP	All	GEN	SP	All	GEN	SP	All	GEN	SP	All	GEN	SP
	(N=500)	(N=250)	(N=250)	(n=100)	(n=50)	(n=50)	(n=100)	(n=50)	(n=50)	(n=100)	(n=50)	(n=50)	(n=100)	(n=50)	(n=50)	(n=100)	(n=50)	(n=50)
I. Inability to communicate	75	68	82*	792	76	82	821	74	90	60	54	66	743	66	82	801	72	88
II. Inability to care for oneself	73	66	79†	67	54	80b	871,2	86	88	62	56	68	69	54	84b	783	82	74
III. Difficulty doing simple tasks like getting dressed	70	61	79*	74	58	90a	63	54	72	73	66	80	69	56	82b	72	72	72
IV. Getting lost or disoriented in a familiar place	57	53	61	791,2	70	88c	58	52	64	44	46	43	47	38	56	58	60	56
V. Bad judgment with common sense decisions	50	45	55‡	573	42	72b	49	44	54	553	56	54	40	36	44	48	46	50
VI. Having safety issues	40	36	43	501	36	64b	571	56	58	27	22	32	22	22	22	432,3	46	40

*Only 6 most frequently chosen responses listed (those with at least 50% frequency from more than 1 group) in the order of frequency chosen by all MDs. GEN vs SP for Total MDs: *P≤0.001; ^†P≤0.01; ^‡P≤0.05. All vs All between countries: ¹P≤0.001; ²P≤0.01; ³P≤0.05. GEN vs SP within countries: ^aP≤0.001; ^bP≤0.01; ^cP≤0.05. All vs All: I. France, Germany, Spain and UK vs Italy. II. Germany¹ vs France and Italy; Germany² vs Spain; UK vs Italy. IV. France1 vs Italy and Spain; France² vs Germany and UK. V. France and Italy vs Spain. VI. France and Germany vs Italy and Spain; UK vs Italy³ and Spain².

Table 2B.

Signs of Early AD

	Frequency item chosen (%)
Response items*	Total			France			Germany			Italy			Spain			UK
	All	GEN	SP	All	GEN	SP	All	GEN	SP	All	GEN	SP	All	GEN	SP	All	GEN	SP
	(N=500)	(N=250)	(N=250)	(n=100)	(n=50)	(n=50)	(N=100)	(N=50)	(N=50)	(n=100)	(n=50)	(n=50)	(N=100)	(N=50)	(N=50)	(n=100)	(n=50)	(n=50)
I. Forgetting a new, recently learned piece of information	72	66	78†	803	68	92b	72	68	76	68	60	76	64	60	68	77	76	78
II. Difficulty in accomplishing tasks at work	63	54	72*	57	44	70c	66	58	74	52	50	68	683	58	78c	66	60	72
III. Difficulty with remembering the time or place	63	62	64	63	68	58	59	58	60	59	48	56	67	62	72	743	74	74
IV. Difficulty with remembering the date	62	62	63	60	60	60	712	64	78	53	58	48	60	54	66	683	72	64
V. Difficulty talking with others, such as forgetting words or substituting unusual words	61	60	63	55	54	56	56	62	50	66	64	68	63	64	62	67	56	78c
VI. Having safety issues	58	55	61	51	54	48	41	36	46	642	64	64	701,2	62	78	642	60	68
VII. Loss of initiative (becoming passive for hours or sleeping more than usual)	54	50	59‡’	56	46	66	56	58	54	47	48	46	643	52	76c	49	44	54
VIII. Rapid mood swings from calm to anger for no apparent reason	54	57	51	532	66c	40	671	74	60	43	42	44	721,2	68	76	34	34	34
IX. Making bad financial decisions	52	45	60*	583	54	62	56	54	58	44	28	60b	622,3	52	72	42	38	46
X. Depression	51	43	59*	58	48	68	56	50	62	44	32	56c	50	38	62c	48	48	48
XI. Getting lost or disoriented in a familiar place	50	50	50	31	32	30	43	46	40	57’	52	62	621,3	68	56	581,3	54	62

All = generalists and specialists combined. *Only 11 most frequently chosen items listed (to capture all responses at or above 50% from all generalists and specialists) in the order of frequency chosen by all MDs; GEN vs SP for Total MDs: *P≤0.001; ^†P≤0.01; ^‡P≤0.05; GEN vs SP within countries: ^aP≤0.001; ^bP≤0.01; ^cP≤0.05. All vs All between countries: ¹P≤0.001; ²P≤0.01; ³P≤0.05. I. France³ vs Spain. II. Spain³ vs Italy. III. UK³ vs Germany and Italy. IV. Germany² vs Italy; UK³ vs Italy. VI. Spain¹ vs Germany; Spain2 vs France; UK² vs Germany. VII. Spain³ vs Italy and UK. VIII. Germany¹ vs Italy and UK; Spain¹ vs Italy and UK; France² vs UK; Spain² vs France. IX. Spain² vs UK; France³ vs UK; Spain3 vs Italy. XI. Italy¹, Spain1 and UK¹ vs France; Spain³ vs Germany; UK³ vs Germany.

The physicians’ appreciation of caregiver concerns, awareness of currently recruiting clinical trials in their area, and knowledge of whom to contact for advice about AD/related dementia were also ascertained.

Practice

Practice-related items included the following: assessment of physician workload, estimation of AD severity at time of presentation, identification of the person who brings the patient to the office for the initial AD-related visit and of the physician who makes the diagnosis and evaluation of treatment and referral goals and practices.

For a more complete description of survey methodology, see the article by Jones et al in this issue (6).

Results

Knowledge of AD

Signs of early and late AD

Physicians seem to be in substantial agreement as to the late signs of AD, with the same 6 signs being chosen most frequently in almost all countries and by both specialists and generalists (Table 2A). Specialists chose 4 of the 5 top responses significantly more frequently than generalists; the discrepancy between specialists and generalists was most pronounced in France, where specialists chose 5 of the 6 top responses significantly more frequently than generalists. In contrast, there were no significant differences between specialists and generalists in Germany, Italy or the United Kingdom, and just 2 in Spain. There were also significant between-country differences for 5 of 6 items, with France choosing 4 of the 6 responses significantly more frequently than at least 1 other country, compared with 3 such responses for Germany, 2 for the United Kingdom and 1 each for Spain and Italy.

There was markedly more variation with respect to the early signs of AD (Table 2B). In order to capture all responses endorsed by at least 50% of generalists and specialists, 11 response items needed to be included. Although these 11 items comprised the top 11 choices in all but a few instances, there remained substantial discrepancies in terms of the frequency with which each item was chosen. For example, significant differences between countries were noted for 9 of these 11 responses, and between generalists and specialists for 5 of 11 responses, including the top 2. Interestingly, responses from specialists and generalists appeared more similar when results within countries were evaluated; there were significant differences for only 3 of these 11 responses in France and Spain, 2 in Italy, 1 in the United Kingdom and none in Germany.

Caregiver concerns

Physicians appeared to be acutely aware of caregiver concerns; each of the 8 response items was endorsed as being at least a minor concern with a frequency of ≥88%. There were virtually no differences between groups in terms of their assessment of caregiver concerns. The largest discrepancy was found for the item “Not knowing if treatments are available to effectively manage the symptoms,” which was viewed as a major concern by 54% of specialists vs 42% of generalists (P<0.05). Physicians from Spain recorded the highest percentage of “major concern” responses for 5 of the 8 items; physicians from Italy recorded the highest percentage of such responses for the other 3 items.

Consultative resources

Almost two thirds of specialists claimed that they “always” know who to contact for AD-related advice, whereas less than half of generalists stated they always know who to contact (P<0.05). However, the survey did not ask whom physicians contacted; the question was open ended, encompassing all aspects of AD care, including medical and mental health treatment and social support and advocacy services. Overall, specialists responded “always” at least 20% more frequently than generalists in every country but France. The range of responses amongst generalists extended from 32% in the United Kingdom to 62% in France; for specialists, responses ranged from 58% in the United Kingdom to 72% in France.

Clinical trials

Just 7% of generalists, and only 32% of specialists (P<0.001), claimed to be aware of an AD clinical trial in their area currently in the process of recruitment. This issue is discussed in much greater detail in the article by Jones et al (14).

Practice

Patient base

Generalists reported seeing twice as many patients per month overall, but relatively fewer geriatric patients and far fewer AD patients than specialists (approximately 1 AD patient out of every 20-25 for generalists compared with 1 of every 5 or 6 patients for specialists) (Table 3). They also spend less than half as much time per patient visit.

Table 3.

Estimate of Patients Seen per Month

	GEN	SP
N (mean, range)	497 (414–636)	237 (188–333)
N (median, range)	500 (400–675)	200 (120–385)
Age ≥65 years (mean, median)	248, 200	149, 100
Patients with AD (mean, median)	27, 20	50, 35
Percentage with AD (mean, median)	5.4, 4.0	21.1, 17.5
Percentage time spent on patient care	92, 95	86, 90
(mean, median)
Total time per patient*	18 minutes	42 minutes

*Based on median percentage direct patient care per 8-hour day and 20 working days per month.

Initial presentation

At the time patients with AD were first seen in a physician’s office for evaluation of cognitive symptoms, physicians as a whole reported that a slim majority of patients (52%) was showing mild symptoms of dementia, and 45% already had moderate symptoms (Fig. 1). Specialists reported that 60% of their AD patients had mild symptoms at the initial visit and generalists reported that just 45% of their AD patients presented with mild symptoms (P<0.001). As expected, a minute proportion of patients had severe symptoms when first seen by either a generalist or a specialist. Spain had the greatest proportion of patients presenting with mild symptoms and Italy the smallest (64% vs 37%, P<0.001).

Figure 1.

Level of AD severity at initial consultation

Diagnosis

Overall, specialists claimed that they diagnose patients with AD themselves in 65% of cases; generalists claim to do so in only 33% of cases (P<0.001, Fig. 2). This difference was of similar magnitude in all countries.

Figure 2.

How AD patients were initially diagnosed

Generalists and specialists indicated overwhelmingly that a family member was most likely to bring the patient in for evaluation (93% overall, 91% generalists and 94% specialists). Germany was the only country in which >10% of initial visits were by the patient alone (12% overall, 14% generalists and 10% specialist).

Referral patterns

The 3 specialties to which generalists most frequently refer their patients for AD-related diagnosis and care are neurologists (64%), geriatricians (40%) and psychogeriatricians (34%). These data are confounded by the between-country differences in specialist physician titles; for example, in the United Kingdom 74% of referrals go to psychogeriatricians, whereas in some other countries this specialty does not exist.

The reason generalists most often cited for why they refer patients was “to begin or ensure treatment” (57%); this reason was cited by just 26% of specialists (P<0.001). In contrast, “ruling out another condition” and “getting a second opinion” were endorsed by similar percentages of generalists and specialists (43% vs 46% and 44% vs 46%, respectively). Interestingly, less than one third of generalists selected “not comfortable making the diagnosis” as a reason for referral, and less than one fifth selected “not my area of expertise.” As expected, specialists did not report referring patients as often as generalists because of a regulatory requirement (14% vs 43%; [P<0.001]). With respect to differences between countries, German physicians endorsed the reason “not comfortable making a diagnosis” most frequently (73% generalists, 39% specialists), while they endorsed the reason “it is a requirement” the least frequently, as expected (7% generalists, 0% specialists). Physicians from the United Kingdom endorsed the reason “to begin or ensure treatment” most frequently (78% generalists, 53% specialists).

Treatment

Goals: The 3 main goals for AD treatment were perceived as slowing memory loss (71%), reducing loss of general functioning (66%) and improving behaviour (66%). Differences between physician groups and between countries in this regard were very small. Not surprisingly, prioritisation of these 3 goals mirrors perfectly the 3 primary outcome measures in AD clinical trials: cognition, function and behaviour.

Treatment initiation: A significantly higher percentage of specialists than generalists initiate treatment overall (94% vs 49%; P<0.001) in every country (Fig. 3). Generalists initiate treatment most frequently in Germany and least in the United Kingdom (78% vs 14%, P<0.001). Amongst physicians who do initiate treatment, a similar percentage of specialists and generalists initiated pharmacotherapy within 1 month of diagnosis (77% and 75%, respectively). The percentage of physicians initiating treatment immediately or within 1 month of diagnosis was consistent across countries (range: 71% in the United Kingdom to 82% in Spain) (Fig. 4).

Figure 3.

Initiation of treatment in AD patients

Figure 4.

Timing between diagnosis and initial prescription treatment

Treatment methods: There was only 1 significant difference in the treatment resources actually used by generalists and specialists: “prescription medication specifically for Alzheimer’s disease,” which was endorsed by 97% of specialists vs 76% of generalists (P<0.001, Table 4). The main prescription treatment options employed overall were AD-specific medication (90%) and medication for mood or behaviour (78%). Not surprisingly, all other treatment options were endorsed at much lower rates, with “regular follow-up” being the only other option endorsed by more than half of all physicians. Most psychosocial interventions were endorsed by around one third of both specialists and generalists.

Table 4.

Treatment Options for Patients With AD

	Frequency item chosen (%)
Treatment options	Total MDs		France		Germany		Italy		Spain		UK
	GEN	SP	GEN*	SP	GEN	SP	GEN*	SP	GEN*	SP	GEN	SP
	(N=123)	(N=235)	(N=20)	(N=49)	(N=39)	(N=49)	(N=25)	(N=49)	(N=32)	(N=49)	(N=7)†	(N=38)
I. Prescription medication specifically for AD	76	971	90	100	92	98	72	94	59	98		95
II. Prescription medication for mood or behaviour	74	81	75	80	69	84	80	71	72	82		90
III. Regular follow-up	56	60	65	63	67	63	44	55	47	48		71c
IV. Daycare	36	40	40	47b	36	41b	20	16	38	42b		61a
V. Care strategies/activities	40	33	45	22	39	31	24	41	50	30		45c
VI. Give caregiver address/telephone number of Alzheimer Patient Association	36	37	55	35	39	37	20	25	28	40		55b
VII. Counseling	39	33	5	29	80	74a	24	18	28	14		32
VIII. Support group	33	37	20	22	46	57a	24	22	22	18		71a
IX. Cognitive therapy	30	35	35	25	36	29	32	43	25	52c		24
X. Psychostimulation,	16	21	40	35a,b	5	4	12	8	19	38a		18
XI. Give caregiver contact information of a key person to contact if in need	17	19	45	10	23	29a,b	4	18c	6	2		40a,b,c
XII. Other prescription medication,	16	14	15	8	5	12	12	16	31	16		16
XIII. Vitamins	13	13	20	12	8	6	12	27b,c	19	4		16
XIV. Psychotherapy,	11	7	15	10	10	6	-	6	16	6		8
XV. Over-the-counter medication,	7	4	10	-	13	8	4	10	3	-		-
XVI. Natural herbs	1	1	-	2	3	-	-	2	-	-		3

*Number of respondents in this category too small for statistical comparisons; †Too few respondents to have meaningful data for this group; GEN vs SP for Total MDs: 1P≤0.001; SP vs SP between countries: ^aP≤0.001; ^bP≤0.01; ^cP≤0.05. III. UK^c vs Spain. IV. UK^a vs Italy; France^b, Germany^b and Spain^b vs Italy. V. UK^c vs France. VI. UK^b vs Italy. VII. Germany^a vs France, Italy, Spain and UK. VIII. Germany^a and UK^a vs France, Italy and Spain. IX. Spain^c vs France and UK. X. France^a vs Germany; France^b vs Italy; Spain^a vs Germany and Italy. XI. Germany^a and UK^a vs Spain; Germany^b and UK^b vs France; Italy^c vs Spain; UK^c vs Italy. XIII. Italy^b vs Spain; Italy^c vs Germany

A comparison of generalists between countries, and of generalists and specialists within countries, was precluded by the limited number of generalist respondents who initiated treatment. Overall, there were significant differences between specialists from different countries for 10 of the 16 treatment options. In particular, there was a significantly higher rate of referral to counseling in Germany, and of referral to support groups in Germany and the United Kingdom, whereas referral to day care was significantly lower in Italy compared with every other country. Modalities specifically addressed toward the caregiver (e.g. providing contact information for an advocacy society) were consistently endorsed by a significantly larger proportion of specialists from the United Kingdom. Cognitive therapy and psychostimulation were recommended most frequently in Spain (52% and 38%, respectively) and vitamins most commonly in Italy (29%).

Discussion

These results describe the similarities and differences between specialists and generalists regarding a wide range of issues pertaining to the diagnosis and management of AD. Perhaps the most important finding was the preponderance of similarities between generalists and specialists. The clinical practices of both groups of physicians demonstrate a strong emphasis on early diagnosis and treatment and prescribing of AD-specific medication, with less consistent referral to a wide range of psychosocial support services. The most obvious, and not unexpected, difference between these 2 groups of physicians is their clinical workload; specialists see fewer patients for longer visits, and also see a greater proportion of geriatric and AD patients. Differences between countries were often more notable than differences between physician groups, most likely as a result of differences in regulatory policy and health care systems (for example, the initiation of AD drugs in the United Kingdom is restricted to specialists) (7). It is not surprising that specialists reported spending more time with each patient and seeing more patients with AD. Although these survey data are not empirical counts of patient visits, they do represent the “best estimate” of the respondents. In no other area of the IMPACT survey was there such a dramatic difference between specialists and generalists. This issue has been raised consistently by generalists when asked to identify barriers to providing more comprehensive care to AD patients and their caregivers (4, 8), and must be realistically addressed if generalists are to optimally fulfill the roles expected of them by patients, caregivers and the health care system (2, 3).

There are a number of limitations of this survey data that should be highlighted. First, the sample was not random; the use of a computer-based format probably skewed participation, and the generalist and specialist physician groups were heterogenous, which may have masked differences within each group. Second, the numbers were small for some comparisons, especially within countries, and there was no correction for multiple comparisons of statistical significance levels. Third, the wording of some questions was ambiguous, resulting in uncertainty in the interpretation of responses. Fourth, the multiple-choice format may have constrained responses to some questions and prevented the acquisition of detailed ad verbatim data.

One of the most striking findings in the survey is that generalists reported that less than half of their patients with AD presented with mild symptoms. This finding provides further evidence suggesting that many patients are not being diagnosed until the moderate stage of their disease (1, 9). Spain appears to be better in this area and may have certain model programmes or policies worthy of emulation (10). Interestingly, specialists reported a higher percentage of their patients presenting with mild AD (60% vs 45% for generalists, P<0.001). One possible explanation for this discrepancy is that specialists are more likely to carry out a detailed assessment of cognitive function whereas generalists are more likely to use a brief cognitive screening test that may not be as accurate at defining the level of cognitive impairment.

Another salient finding is that just one third of generalists reported diagnosing their patients with AD, which was complemented almost exactly by specialists reporting that they diagnose two thirds of their AD patients. However, these results must be considered in light of existing regulations/best practice guidance in most of the countries surveyed that require diagnosis of AD by a specialist in order to commence antidementia treatment. Germany is an exception in this regard, and France also allows a subset of generalists to treat directly as well. Nonetheless, generalists may still diagnose patients with AD despite this regulatory requirement concerning initiation of treatment.

Physicians from all countries and both groups (specialists and generalists) identified the late symptoms of AD with a high degree of consistency but demonstrated substantial variability with regard to the early symptoms of AD, revealing differences across countries and, to a lesser extent, between specialists and generalists within each country. This lack of consistency in knowledge of early symptoms of AD reflects the variation in symptoms with which people in early phases of dementia present in clinical practice; including cognitive and non-cognitive difficulties, such as speech, mood and personality changes. This may contribute to variability in the accuracy and timing of diagnosis and most likely contributes to the perception, particularly amongst caregivers, that generalists and specialists both have difficulty recognising early AD (see the article by Wortmann et al (11)). Although, as one would in this issue expect, specialists performed better than generalists (i.e. their most frequent responses adhered more closely to cognitive, functional and behavioural signs of mild AD), the differences between the groups were not large, suggesting that generalists may be capable of making the diagnosis almost as well as specialists (12, 13).

One of the most consistent findings across physician groups and countries was that almost all patients are brought to the physician initially by a family member. Although geriatric patients may be somewhat more likely than other adult patients to have a family member accompany them to the doctor, it is unlikely that this pattern is observed as dramatically for any other disease. Although not surprising, this result should strengthen the argument that any programme whose goal is to decrease the delay to diagnosis of AD must be designed to engage family members at least as much as patients (14).

Physicians tended to recognize caregiver concerns accurately, and identified most of the options offered in the survey as being of major concern to caregivers. Overall, there were remarkably few differences between physician groups and between countries. The 2 items that were chosen least frequently as major concerns may have been so because the wording of the items made them seem to be more of an observation of the caregiver by the physician than a concern of the caregiver per se (“Not recognising the seriousness of the symptoms” and “being in denial about the implications of the symptoms”). Interestingly, almost half of physicians identified uncertainty as to the availability of treatment as a major concern, suggesting that greater efforts could be made to make the public aware of treatment options for AD. Doing so may lessen the intensity of the fear of the disease and promote earlier recognition and treatment-seeking.

Specialists felt more confident about their consultative resources (i.e. were more often sure about whom to consult for AD-related matters) than generalists did; this trend was large and consistent across Europe. However, given that patients and caregivers are at least as likely to seek assistance and care from their generalists, it would seem important to increase the ability of these physicians, or their office staff, to offer appropriate and relevant information in a timely manner.

Physicians’ treatment goals were nearly identical in all countries and amongst specialists and generalists. The major difference between physician groups was in terms of treatment initiation; twice as many specialists as generalists reported initiating therapy. This most likely reflects regulatory requirements that prohibit generalists from initiating treatment in some countries (e.g. United Kingdom). Although it is encouraging that roughly three quarters of physicians who initiate treatment do so either immediately or within 1 month of diagnosis, these results indicate that about one quarter of physicians wait 2 months or longer after diagnosis before initiating therapy. As clinical trial data have shown, delayed initiation of therapy for AD can have long-term negative effects on disease progression (15).

A second notable difference between generalists and specialists was the prevalence of prescribing AD-specific medication, which was almost universal amongst specialists but reported by just three quarters of generalists. Although this, too, may ultimately reflect regulatory policies, it may also be a result of specialists’ perceiving more benefit for these medications in later stages of AD (see the article in this issue by Martinez-Lage et al (16)). In other respects, the types of therapy employed by generalists and specialists did not differ remarkably. With few exceptions (e.g. counseling in Germany, cognitive therapy in Spain, support groups in the United Kingdom), rates of referral to psychosocial treatment and support services were relatively low, suggesting that physicians may not be utilizing these modalities optimally despite their proven benefits for both people with dementia and their carers (3). In addition to providing patients a timely, accurate diagnosis and appropriate pharmacological therapy, it appears that physicians could play a larger role in directing patients and caregivers to psychosocial and other support services from which they would also benefit, though the availability of such services in some countries may be limited (3).

Perhaps the most interesting result in the area of referral patterns was the relatively small percentage of generalists who reported referring patients because they were not comfortable making the diagnosis or because AD was not their area of expertise. The major exception to this pattern was Germany, where a very high percentage of both generalists and specialists reported they were uncomfortable making the diagnosis. Generalists cited government regulations as a common reason for making a referral, especially in the United Kingdom. These data indicate that a sizable majority of generalists believe they have sufficient knowledge and skills to diagnose and treat patients with AD and are comfortable deciding when they need specialist input.

In conclusion, there were more similarities than differences amongst generalists and specialists regarding a broad spectrum of issues relating to AD. Where differences occurred, they were as likely to be between countries, perhaps secondary to regulatory requirements, as between the physician groups. Where physician groups differed, specialists displayed greater knowledge, and their practices demonstrated a more intensive treatment orientation, as would be expected. These results show that both generalists and specialists are actively involved in the diagnosis and management of AD. There are areas, however, such as referring for psychosocial interventions and other support services, and having knowledge of research opportunities, where both groups of physicians need to be better informed in order to improve the care of their patients with AD.

Disclosure: The IMPACT study was funded by Pfizer Inc and Eisai Inc. This article was funded by Pfizer Inc and Eisai Inc. Editorial support was provided by Bill Kadish, MD, of PAREXEL and was funded by Pfizer Inc and Eisai Inc.