The integration of dermatology experts into primary care to assess and treat patients with skin lesions is cost‐effective: A quasi‐experimental study

Abstract

Background

The management of patients with skin changes can be challenging in primary healthcare; general practitioners (GPs) often lack the expertise to make accurate assessments and treatment decisions. The standard care pathway for skin changes can result in extended treatment times and costs.

Objectives

This study was designed to evaluate the cost‐effectiveness of integrating a dermatologist into the primary care setting to assess and treat patients with skin disorders. The primary outcome was the incremental cost‐effectiveness ratio (ICER) for each malignant or pre‐malignant skin disease found and treated. The secondary outcomes included ICER for any treated skin finding, number needed to excise to find malignant or pre‐malignant skin disease, number of hospital referrals required and changes in quality of life (QoL) in the presence and absence of the integration.

Methods

This was a quasi‐experimental cohort study conducted at three primary healthcare centres in Finland. In the two intervention centres, patients with skin findings visited a dermatologist; in the control centre they visited a GP. Cost‐effectiveness was assessed using the incremental cost‐effectiveness ratio (ICER). QoL was assessed with the PROMIS v1.2, calculative EQ‐5D‐3L and PROMIS Anxiety 4a instruments.

Results

In total, 186 integration and 176 control patients were included. For an additional patient treated for a (pre‐)malignant skin disease, the ICER was €852 lower and with any skin disease €381 lower in the integration group than with standard care. Fewer biopsies were required for each malignant or pre‐malignant skin disease in the integration group compared to the control group (2.1 and 6.5 per patient; p < 0.001) and lower proportion of patients were referred to hospital (8.1 vs. 17.1%, p < 0.001). Patient QoL did not differ between groups.

Conclusions

The integration of dermatological expertise into primary care settings is cost‐effective and can streamline the management of patients with skin conditions without worsening their QoL.

The integration of dermatologist into primary care settings is cost‐effective and streamlines the management of patients with skin conditions. The cost for each additional pre‐malignancy or malignancy treated is €853 lower, and for any additional skin finding treated, €381 lower than that of the standard care pathway.

Why was this study undertaken?

• Many skin changes, even cancers, remain undetected at general practitioner appointments. The cost‐effectiveness of adding dermatologists to primary care teams has not been widely investigated.

What does this study add?

• When a dermatologist works in primary care, the cost for each additional malignancy or pre‐malignancy found and treated is €853 lower than that of the standard care pathway, while the cost for any additional treated skin finding is €381 lower.

What are the implications of this study for disease understanding and/or clinical care?

• The integration of a dermatologist into the primary care setting is favourable, in terms of both medical outcomes and cost‐effectiveness.

INTRODUCTION

Patients with skin findings frequently require biopsies, which impose a burden on the healthcare system at various levels. ¹ Many signs of skin cancer escape detection at general practitioner (GP) appointments, ² , ³ while the potential for malignancy causes distress to patients with skin conditions. ⁴ There are indications that the addition of a dermatologist to the primary care team reduces the number of patients needed to excise (NNE) to detect melanoma. ⁵

Dermatological disorders are among the most common diseases detected at GP appointments. ⁶ , ⁷ , ⁸ , ⁹ Over 70% of the population aged over 70 years have at least one skin disease diagnosis, as do over 60% of a middle‐aged people; nearly half of them require further treatment. ¹⁰ , ¹¹ Non‐melanoma skin cancer including basal cell carcinoma (BCC) and squamous cell carcinoma is the most common malignancy in people with fair skin, ¹² worldwide, over one‐third of carcinomas are skin carcinomas. ¹³ Since the 1960s, the incidence of skin cancer has increased ninefold according to the Finnish cancer register. ¹⁴

We recently performed a systematic literature review, which found that the evidence for the cost‐effectiveness of placing hospital specialists in primary care is inconclusive. ¹⁵ The cost‐effectiveness has not been extensively researched with regard to the placement of dermatologists in primary care. ¹⁵ , ¹⁶ , ¹⁷

The objective of this intervention study was to explore whether the early involvement of a dermatologist in the primary healthcare setting could improve the cost‐effectiveness of the care pathway for patients with skin findings, while considering the perspectives of both patients and commissioners.

MATERIALS AND METHODS

This was an empirical multicentre quasi‐experimental observational study of a new model of primary care (subsequently referred to as the ‘intervention’) for patients with skin disorders. Patients at two primary healthcare centres in northern Finland were assigned to the intervention group, while patients attending a third centre were assigned to the control group. In the intervention group, the patients' primary care doctor meeting was with a dermatologist, without need to meet a general practitioner (GP) first, while patients in the control group received standard care and were initially seen by a GP. Patients were followed‐up until their skin disorder was treated to a maximum of 1 year.

The patient study population was recruited from 2021 to 2022 among all the skin finding patients over 18 years who either contacted a health centre with a skin finding or on whom a GP or a nurse noticed a skin finding which needed to be assessed. The skin changes like melanocyte naevi, lumps and unknown or non‐healing eczemas or scab were prioritized. Participants were required to provide written informed consent voluntarily and comply with the study protocols. Patients currently undergoing treatment for the same skin lesion were excluded from the study. In the study, the diagnosis was considered as confirmed if it was made based on histopathological analysis (PAD) or dermoscopy. Histopathological analysis was prepared in the Central Hospital in Kemi.

The healthcare system in Finland is publicly funded and operates on a capitation basis. Each health centre provides primary healthcare to all the residents in their catchment area.

Data collection

Prior to the initial appointment, patients completed a structured quality of life questionnaires PROMIS v1.2 ¹⁸ and PROMIS 4a ¹⁹ in paper format and a follow‐up subsequently, after 6–12 months. Clinical parameters, including diagnoses and treatment given, previous skin diseases, multimorbidity and smoking status, were collected from the electronic patient record system OMNI by CGI. ²⁰ Multimorbidity was defined a status where a patient has at least three long‐term illness for which he/she receives regular care or is monitored by the healthcare professional. Long‐term refers to a period of at least 6 months for which the illness has continued or is expected to continue. ²¹

Outcomes

The primary study endpoint was the cost‐effectiveness of the care chain for management of malignant or pre‐malignant skin conditions, as measured by the incremental cost‐effectiveness ratio (ICER), calculated as shown below:

$$\text{ICER}=\frac{\text{Costintervention care chain}-\text{Costcontrol care chain}}{\text{Patients with}\left(\mathrm{Pre}\right)\text{malignitiesintervention care chain}-\text{Patients with}\left(\mathrm{Pre}\right)\text{malignitiescontrol care chain}}$$

The secondary outcomes were ICER for any detected skin disease, the total cost of the care chain, QoL, measured by the PROMIS v1.2 instrument, of which also estimated EQ‐5D‐3L was derived, ¹⁸ and PROMIS Anxiety 4a instrument. The clinical/process outcomes measured were the number of biopsies performed for each malignant or pre‐malignant disease diagnosed, total numbers of patients diagnosed with malignancy, pre‐malignancy, or only seborrhoeic keratosis, numbers of patients biopsied, cryo‐treated in primary care, or prescribed a skin medicine and number of patients referred to the hospital.

Costs

The actual unit costs associated with labour in primary care were meticulously computed for the study health centres. The costs for each centre pertaining to infrastructure and administrative overheads were determined based on the genuine expenditures incurred by one of the study centres. To calculate the labour cost of care for each patient, the unit cost was multiplied by the time expended by each healthcare professional throughout the patient's care pathway. Diagnosis‐related group (DRG) pricing was applied to determine the cost of hospital visits. A unit cost of €80 was used for each histopathological analysis performed, which included all expenses and overheads.

The intervention required the provision of a dermoscope and a pot for liquid nitrogen, for which a cost of €2.80 was applied for each patient in the integrated care chain, under the assumptions that 70% of the use of the equipment would be for investigation and treatment of the intervention population and that the value of each item would fully depreciate over a period of 4 years.

The travel cost for the dermatologist €19.90 per day in primary care, based on the data from the Finnish Transport Infrastructure Agency ²² and the Social Insurance Institution of Finland, ²³ as well as the cost of their travel time (60 min per intervention day) were included in the overall unit price of the specialist.

Cost calculation cover each patients' care pathway to a maximum of 12 months after the first contact. Statistical analysis of costs was performed only for commissioner costs.

Statistical analysis

According to the power calculation, 140 patients were needed per group to obtain a power of 80% (Fisher's exact, two‐sided test, α = 005). ²⁴ Categorical variables were tested with Pearson's chi‐squared test or Fisher's exact test and continuous variables with the Mann–Whitney U‐test. Logistic regression was applied to the categorical variables and a linear regression model for the continuous variables to estimate the associations between the intervention and dependent variables. Crude and adjusted odds ratios and confidence intervals with p‐values were determined. Age, sex, smoking status, previous skin diseases and multimorbidity were adjusted for as confounders. Missing data were not imputed in the analyses. p‐Values <0.05 were considered statistically significant.

The R studio software package ²⁵ version 4.2.2 was applied for the statistical analysis of the research data.

RESULTS

Baseline status

The study included 362 individuals with a skin finding, 186 in the intervention group and 176 in the control group. A history of previous skin disease was more common in the intervention group than the control group (42.2% vs. 28.1%, p = 0.005) but there were no differences in age, sex, smoking or multimorbidity. The intervention and control groups were comparable at baseline in all measured QoL parameters except for estimated EQ‐5D‐3L score, which was significantly higher in the intervention group (Table 1).

TABLE 1.

Baseline characteristics.

Parameter	Intervention (N = 186)	Control (N = 176)	Total (N = 362)	p‐Value
Gender				0.558
Male	57 (30.6%)	59 (33.5%)	116 (32.0%)
Female	129 (69.4%)	117 (66.5%)	246 (68.0%)
Age				0.789
Median (IQR)	66 (47, 74)	65 (45, 73)	65 (46, 74)
Range	19–93	18–96	18–96
Mean (SD)	60 (18)	60 (19)	60 (18.3)
N	186	176	362
Smoking				0.118
No	105 (68.2%)	90 (57.0%)	195 (62.5%)
Ex	33 (21.4%)	44 (27.8%)	77 (24.7%)
Yes	16 (10.4%)	24 (15.2%)	40 (12.8%)
Previous skin diseases				0.005
No	106 (57.6%)	123 (71.9%)	229 (64.5%)
Yes	78 (42.4%)	48 (28.1%)	126 (35.5%)
Multimorbidity a				0.582
No	93 (50.3%)	92 (53.2%)	185 (51.7%)
Yes	92 (49.7%)	81 (46.8%)	173 (48.3%)
PROMIS Global Health v1.2
Global Physical Health T‐score				0.138
N	177	149	326
Median (IQR)	47.7 (42.3, 54.1)	47.7 (39.8, 50.8)	47.7 (42.3, 54.1)
Mean (SD)	48.2 (8.0)	46.8 (8.9)	47.6 (8.5)
Range	29.6–61.9	23.5–67.7	23.5–67.7
Global Mental Health T‐score				0.696
N	178	148	326
Median (IQR)	48.3 (43.5, 53.3)	48.3 (43.5, 53.3)	48.3 (43.5, 53.3)
Mean (SD)	48.3 (7.8)	48.5 (8.4)	48.4 (8.0)
Range	28.4–67.6	25.1–67.6	25.1–67.6
Self‐rated health	185	171	356	0.658
Poor	3 (1.6%)	5 (2.9%)	8 (2.2%)
Fair	45 (24.3%)	44 (25.7%)	89 (25.0%)
Good	62 (33.5%)	61 (35.7%)	123 (34.6%)
Very good	64 (34.6%)	48 (28.1%)	112 (31.5%)
Excellent	11 (5.9%)	13 (7.6%)	24 (6.7%)
Ability to carry out social activities and roles	185	170	355	0.411
Poor	1 (0.5%)	4 (2.4%)	5 (1.4%)
Fair	22 (11.9%)	16 (9.4%)	38 (10.7%)
Good	59 (31.9%)	59 (34.7%)	118 (33.2%)
Very good	74 (40.0%)	59 (34.7%)	133 (37.5%)
Excellent	29 (15.7%)	32 (18.8%)	61 (17.2%)
PROMIS Adult Short Form v1.0—Anxiety 4a
Anxiety T‐score				0.759
N	170	167	337
Median (IQR)	51.2 (40.3, 55.8)	51.2 (40.3, 57.7)	51.2 (40.3, 55.8)
Mean (SD)	50.5 (8.5)	50.9 (8.5)	50.7 (8.5)
Range	9.0–73.3	40.3–71.2	9.0–73.3
EQ‐5D‐3L, estimated				0.004
N	177	149	326
Median (IQR)	0.72 (0.64, 0.78)	0.69 (0.58, 0.76)	0.71 (0.62, 0.78)
Mean (SD)	0.71 (0.1)	0.66 (0.1)	0.68 (0.1)
Range	0.19–0.87	0.19–0.88	0.19–0.88

Abbreviations: IQR, interquartile range; SD, standard deviation.

^a Multimorbidity = minimum three illnesses for which he/she receives regular care or is monitored by the healthcare professional for a period of at least 6 months.

Cost‐effectiveness findings

Compared with the control group, the commissioner ICER was – €852 for each additional patient in the intervention group found and treated for malignant or pre‐malignant skin disease and ‐€381 for each additional patient treated for a skin disease of any type (Table 2).

TABLE 2.

Cost‐effectiveness ratio for intervention and control group patients.

	Intervention, €	Control, €	ICER a , €
Commissioner total cost for the care chain	40,376	54,858
Patients treated for malignancy b	35	18	−852
Patients treated for any skin problem	97	59	−381

^a Each patient was only counted once, even if they presented with, and were treated for, multiple skin findings at the same visits.

^b Includes pre‐malignant disorders; ICER, incremental cost‐effectiveness ratio.

Table 3 shows the comparative costs in the intervention and control groups prior to adjustment. The average total cost of the whole care chain of each patient was significantly lower in the intervention group than in the control group. There was no difference between the groups in the cost of nurse work, while the cost of primary care doctor work, the cost of histopathological exam (pathological anatomic diagnosis, PAD), total primary care cost and hospital cost were all significantly lower in the intervention group. The median hospital cost was zero in both groups because only a minority of patients required hospital referral.

TABLE 3.

Costs in each group.

Cost	Intervention (N = 186)	Control (N = 176)	Total (N = 362)	p‐Value
Total cost				<0.001
N	186	176	362
Median (IQR), €	100.3 (84.6, 282.2)	219.3 (131.4, 321.3)	172.5 (97.5, 299.6)
Mean (SD), €	217.1 (257.5)	311.7 (318.7)	263.1 (292.3)
Range, €	15.3–1877.1	43.7–2567.9	15.3–2567.9
Nurse work cost				0.612
N	186	176	362
Median (IQR), €	53.9 (38.6, 70.7)	50.8 (38.6, 72.2)	51.4 (38.6, 72.0)
Mean (SD), €	58.5 (24.3)	60.2 (38.0)	59.3 (31.6)
Range, €	15.3–170.9	0.0–334.6	0.0–334.6
Primary care doctor work cost				<0.001
N	186	176	362
Median (IQR), €	45.9 (45.9, 125.1)	102.0 (80.5, 161.1)	80.5 (45.9, 141.0)
Mean (SD), €	83.6 (51.8)	122.0 (52.7)	102.3 (55.6)
Range, €	0.0–268.8	0.0–290.0	0.0–290.0
PAD cost				<0.001
N	186	176	362
Median (IQR), €	0.0 (0.0, 80.0)	80.0 (0.0, 80.0)	0.0 (0.0, 80.0)
Mean (SD), €	31.8 (54.1)	52.3 (68.4)	41.8 (62.2)
Range, €	0.0–320.0	0.0–560.0	0.0–560.0
Primary care total cost				<0.001
N	186	176	362
Median (IQR), €	99.8 (84.6, 275.5)	213.4 (130.9, 298.7)	165.2 (96.9, 281.7)
Mean (SD), €	173.9 (119.6)	234.5 (126.6)	203.3 (126.6)
Range, €	15.3–599.6	43.7–970.5	15.3–970.5
Hospital cost				0.006
N	186	176	362
Median (IQR), €	0.0 (0.0, 0.0)	0.0 (0.0, 0.0)	0.0 (0.0, 0.0)
Mean (SD), €	43.2 (194.7)	77.2 (255.5)	59.7 (226.6)
Range, €	0.0–1573.0	0.0–1975.0	0.0–1975.0

Abbreviations: IQR, interquartile range; SD, standard deviation.

After adjustment in the linear regression analysis, the effect of the intervention on costs for the commissioner remained significant for the total care chain cost and all the cost sub‐categories apart from nurse cost and hospital cost (Table 4). Costs incurred by patients were not included in the ICER or cost calculations presented above; however, excess travel cost caused by the hospital visit is presented in Table S1.

TABLE 4.

Linear model statistics for the isolated effect of the intervention on costs for the commissioner.

€	Estimate	Estimate	Std. error	Std. error	Statistic	Statistic	p‐Value	P‐Value
	Crude	Adjusted	Crude	Adjusted	Crude	Adjusted	Crude	Adjusted a
Nurse cost	1.8	2.4	3.3	3.4	0.5	0.7	0.6	0.5
Doctor cost	38.4	38.9	5.5	5.6	7.0	7.0	<0.005	<0.005
PAD cost	20.4	20.5	6.5	6.6	3.2	3.1	<0.005	<0.005
Primary care cost	60.6	61.7	12.9	13.1	4.7	4.7	<0.005	<0.005
Hospital cost	34.0	42.3	23.8	24.2	1.4	1.7	0.153	0.081
Total cost	94.6	104.0	30.4	30.7	3.1	4.1	<0.005	<0.005

^a Adjusted for sex, age, smoking, multimorbidity and previous skin diseases.

Clinical and process outcomes

The number of diagnoses per visit was greater in the intervention group compared to the control group (2.8 and 1.1, respectively; p < 0.005). Fewer biopsies were taken in the intervention group for each detected case of malignant or pre‐malignant skin disease compared with the control group (2.1 and 6.5, respectively; p < 0.005). The distribution of the PAD‐confirmed diagnoses is presented in the Table S2.

The findings of the logistic regression model for the categorical clinical and process endpoints are shown in Figure 1. Patients in the intervention group had significantly lower odds ratio (OR) to require a biopsy, but significantly higher OR to have pre‐malignant or malignant disease detected, although the difference was not significant with regard to malignancies only. Diagnoses of seborrheic keratosis and hospital referrals were more common in the control group, while cryotherapy and pharmacotherapy in the primary care setting were far more likely with the intervention. The frequencies of unadjusted clinical and process parameters with p‐values are presented in Table S3.

FIGURE 1.

Adjusted odds ratios for the categorical process variables.

During follow‐up, there were no differences between groups in measures of QoL or anxiety (Table S4). Within‐group analyses comparing baseline with post‐intervention follow‐up scores we found that the global physical score (PROMIS v1.2) increased significantly from baseline in the intervention group, as did the estimated EQ‐5D‐3L in both groups. The anxiety (PROMIS 4a) fell from baseline in the control population. No other significant changes were recorded in QoL measures inside the groups (Table S5).

DISCUSSION

Managing patients with skin changes in primary care is a multifaceted challenge that heavily draws on healthcare resources, while the capacity to assess skin findings may be suboptimal. ² We assessed the cost‐effectiveness of a new care pathway, in which patients with skin findings are evaluated in primary care by a dermatologist rather than a GP.

Based on the ICER calculated from our study data, it is cost‐effective to bring a dermatologist into primary care. Numerous studies and reviews have investigated the cost‐effectiveness of treating single dermatological diseases, such as melanoma, ²⁶ BCC ²⁷ or hidradenitis suppurativa. ²⁸ However, literature is scarce regarding the optimization of dermatological care in a primary healthcare population, with a cost‐effectiveness focus.

Previous studies have reported various effects on costs of introducing specialists to the primary healthcare pathway. A study from 2008 found that such a scenario reduced the cost per patient by 52% compared with a hospital outpatient clinic pathway ²⁹ and in a newly published Dutch study; the appointment cost per patient was reduced by €10 during a 1‐year programme of specialist involvement in primary care. ³⁰ Outreach clinics in the United Kingdom in 1997 reported significantly lower health service total cost per patient than hospital outpatient care, albeit with a higher marginal cost. ³¹ , ³² Our study demonstrated that, after adjusting for confounders, the intervention reduced the overall cost of the care pathway. We found that the most substantial contributor to cost reduction was the saving of physician time in primary care, followed by the reduction in primary care PADs. The cost of nurse time remained unchanged because the specialist required two assistants or nurses to optimize their own time. The cost‐effectiveness of care model is sensitive to the unit costs of doctors. In our case the unit cost of the dermatologist, even though a senior consultant, was not more than 10% higher than that of a GP.

Service providers have previously highlighted the increased costs and challenges associated with equipping multiple sites for specialist care, as well as the loss of clinical time for the specialists, due to the need for them to travel between locations. ²⁹ In our research, simple investments like a dermoscope, did not play a significant role in cost‐efficiency. As the specialist or GP is the most expensive part of the care pathway, their time needs to be used wisely—days need to be well organized and long enough to avoid prohibitively high relative travelling costs, including time spent travelling.

In our intervention group, the combined number of malignancies and pre‐malignant conditions detected and treated in primary care was double that seen in the control group, even though there was significantly less time spent by the specialist per patient. The hit rate of the biopsy was over threefold greater in the intervention group than in control group. Elevated rates of precision for diagnosis of melanoma and other conditions in outreach clinics have also been reported in some previous studies. ²⁶ , ³³

We found no differences in QoL measures between the intervention and control groups. This outcome was perhaps unsurprising, given that the treatment provided to patients remained consistent, albeit with a reorganization of resources. Furthermore, the impact of predominantly benign skin findings may not be reflected in assessments of overall health. The number of patients answering to follow‐up was also limited; however, there was no differences between the groups (in age, sex, smoking, multimorbidity or previous skin diseases). We did not employ skin symptom‐specific QoL measures, because of the lack of a comprehensive scale covering several skin symptoms.

There were significantly fewer hospital referrals in our intervention group than in the control group. Similar findings were yielded by the above‐mentioned Dutch study, in which a dermatologist was made available to advise GPs, and by an Israeli military programme that placed senior civilian specialists from several disciplines to work alongside military physicians. ³⁰ , ³⁴ Clinical skills, seniority and a certain mindset are crucial qualities for any specialist entering the primary care setting. ³⁵ , ³⁶

The previous studies we found differed from our own methodology in that, as far as was reported, they scheduled a GP appointment before the initial dermatologist meeting, and the role of the dermatologist was mainly consultative, as opposed to our direct replacement.

Considering alternatives for an integrated dermatologist, artificial intelligence (AI) has demonstrated promising results as a future option for the early identification of suspicious pigmented skin lesions in primary care. ³⁷ , ³⁸ A significant barrier to the widespread adoption of AI tools might be as yet limited availability of reliable and integrated systems. ³⁹ Additional research is needed to verify the safety and cost‐effectiveness. ³⁷ , ³⁸ , ⁴⁰ In subsequent years, GPs with special education in dermatology (GPwSE) assisted by the AI tools could potentially partly replace the visiting dermatologist, if proven cost‐effective. Without AI tools GPwSEs may incur greater expenses than hospital outpatient care ⁴¹ and maintaining their skills might be challenging due to workload and diverse patient populations. Teledermatology, recognized as a valid method, ³⁹ , ⁴² , ⁴³ still faces the challenge of ensuring consultations for relevant skin lesions. ²

Based on our experience, the main challenges for a care model similar to our intervention lie in the comparative vulnerability of its set‐up: if the specialist were to call in sick, a substitute may not be available the same day. However, a similar scenario could equally arise in the setting of a small hospital. The availability of rooms could present difficulties in some centres: in our experience it is advisable to have at least two rooms allocated to the dermatologist and their assistant(s) for a smooth process. Additionally, it may be efficient to have another doctor conducting biopsies among his other work while the dermatologist performs her tasks to minimize delays, excess work and patient travel costs.

Although our population was sufficiently large to provide statistical power, a larger sample may have offered more robust results. In terms of study design, a randomized controlled trial (RCT) could have provided stronger internal validity but might have been more complicated and expensive to perform, and would not have provided evidence of real‐world effectiveness. ⁴⁴ Compared to an RCT, an observational practice‐based study such as ours may offer a higher degree of external validity; findings of such studies can be more generalizable and suitable for translation into real‐world practice. ⁴⁵ , ⁴⁶ One limitation in our study is the absence of data regarding the patients who refused to participate. According to the report of the assisting nurses and doctors' daily lists, very few patients declined, unlikely to influence the results.

The presented care pathway could be applied across various healthcare systems. However, while earlier detection of malignancies and fewer referrals could eventually offset the increased demand for primary dermatology resources, the potential shortage of dermatologist might complicate its implementation at the outset. ⁴⁷ , ⁴⁸ Therefore, a collective discussion and consensus on the optimal allocation of resources is necessary.

A strength of our study is the use of the actual unit costs and time spent by the different professionals to determine the true cost of the care pathway in the real‐world primary care setting, and the calculation of the cost per clinical outcomes. Our study appears to be the first to assess the effects of directly replacing the GP with a dermatologist in primary care for the evaluation of patients with certain dermatological conditions and is unique in measuring the cost‐effectiveness of the care pathway for patients with skin conditions.

In summary, patients with skin disorders are a large and growing group that place an increasing demand on primary care resources. Most primary care centres need to assess the needs of all patients who present, and the overarching question in primary care is how to ensure patients get the most benefit—including which professionals to allocate and at which stage of the care chain.

Our findings show that integrating a dermatologist into the primary care setting could streamline the management of these patients and reduce the need for hospital referrals, thus improving cost‐effectiveness. According to our experience, some dermatological conditions, such as acne, rosacea and some dermatitis and inflammatory findings could still be appropriately recognized and treated by a GP. This echoes the findings of some previous studies. ⁴⁹ , ⁵⁰ Further research is required to clarify the conditions under which a patient should be sent directly to a primary care dermatologist and to examine the implications of adding consultative time to the dermatologist's day.

AUTHOR CONTRIBUTIONS

Maria Lovén: conceptualization, data curation and analysis, writing—original draft preparation and revision; Laura Huilaja: supervision and writing; Markus Paananen: supervision; Paulus Torkki: conceptualization, methodology and supervision.

FUNDING INFORMATION

The study was partly funded by grants from Northern Finland Health Care Support Foundation ‐ Terttu Säätiö and Aarne and Aili Turunen Foundation.

CONFLICT OF INTEREST STATEMENT

The authors have no conflict of interest to declare.

ETHICAL APPROVAL

Reviewed and approved by the Medical Ethics Committee of the Northern Ostrobothnia Hospital District (reference EETTMK: 6/2021). The study was conducted according to the principles of the Declaration of Helsinki.

ETHICS STATEMENT

The participants took part on a voluntary basis and provided signed informed consent. The data were handled at the group level—complete participant anonymity was maintained by the replacement of personal information with identification codes.

Supporting information

Tables S1–S5

Lovén M, Huilaja L, Paananen M, Torkki P. The integration of dermatology experts into primary care to assess and treat patients with skin lesions is cost‐effective: A quasi‐experimental study. J Eur Acad Dermatol Venereol. 2025;39:1666–1674. 10.1111/jdv.20451

DATA AVAILABILITY STATEMENT

Upon reasonable request, the corresponding author will share the dataset description to enable the replication of data extraction from healthcare providers' databases.