Multiple Sclerosis Nursing to Improve Care and Education (MSNICE): an observational study

Multiple sclerosis (MS) is an inflammatory and degenerative disease of the central nervous system [1, 2], with an average age of onset between 20 and 40 years. In Belgium, MS affects approximately one in thousand persons with a predominance in women (3:1) [3]. The variety in symptoms, heterogeneity in disease course [4] and monitoring requirements of immunomodulatory treatments make MS a complex disease that requires a multidisciplinary approach [5, 6]. Within the multiple sclerosis care unit, the specialized neurologist and MS-nurse play a central role [7] in the care for persons with MS (pwMS). However, evidence that multidisciplinary or multiple sclerosis-nurse (MS-nurse) led care improves outcomes for pwMS is only emerging and the knowledge gap concerning the added value of an MS-nurse remains large. According to the Multiple Sclerosis Trust, an MS-nurse is considered to be the point of contact, who provides information and gives clinical support. In several countries, such as the United Kingdom [8], MS-nurse led care is a fixed value. In Belgium, several initiatives have been set up, yet to this date, there is no clear job description nor standard implementation of an MS-nurse, which may cause a variety of the given care. Firstly, this explorative study tried to look at differences in centers with or without an MS-nurse regarding health related quality of life (HRQoL), care-related outcomes and educational differences by the means of patient reported outcomes (PROs) covering HRQoL and comorbidities; and health care resource use and expenditures of pwMS. Secondly, the determinants of HRQoL were investigated.

A multicenter, observational and cross-sectional study assessing the impact of MS-nurse care on patient reported outcomes, health care resources use and expenditures in pwMS who had access to an MS-nurse in comparison to those who did not. The following demographics were collected: age, gender, marital status, socio-economic status and education. In order to describe the population, the following variables were collected: Expanded Disability Status Scale (EDSS), number of relapses during the last year, current and previous disease modifying treatments, year of disease onset and year of diagnosis. To compare disease specific quality of life, the multiple sclerosis impact scale-29 (MSIS-29) total, psychological and physical was used as primary outcome measure. The secondary outcome measures were: the Short Form-36 Health Survey (SF36), EuroQoL-5D-5 L (Eq. 5D), Hospital Anxiety and Depression Scale (HADS, 0–7 = normal, 8–10 borderline, abnormal), MS-knowledge questionnaire (inhouse translation, scores 0–25), Coping Inventory for Stressful Situation (CISS), Patient Determined Disease Steps (PDDS). For health care resource use, the number of general practitioner visits related to multiple sclerosis, number of neurological visits, professional help (home care, home nursing), medication use, physiotherapy use, hospitalizations were assessed. PwMS were included if they had a definite diagnosis of relapsing remitting, secondary or primary progressive MS. For centers with an MS-nurse, the condition was that the participant needed to have had at least one contact with the MS-nurse before inclusion. PwMS were excluded for the following reasons: (i) The pwMS was seen as a second opinion, and had attended one of the other participating centers before; (ii) presence of other systemic diseases or serious psychiatric disorders (bipolar disorder, psychosis, etc.) that could interfere with patient reported outcomes, including quality of life measurements; (iii) cognitive dysfunction, as judged by the investigator, to be too severe to complete the questionnaires. Depression was not considered as an exclusion criterion, as we hypothesized that depression may improve in pwMS attended by an MS-nurse. The study was initiated at the Antwerp University Hospital on 11th of August 2014 after receiving ethics committee approval (Belgian registration number B300201421282) and the first patient was included on August 12th 2014. Initially twelve centers (hospitals) were included but, due to insufficient and slow recruitment, after an amendment in October 2019, seven more centers were added. In seven out of the nineteen general or university hospitals, an MS-nurse was available. There were no predefined standards or criteria for the participating MS-nurses. The participating centers were chosen based on the availability of a neurologist with a special interest in MS. The inclusion stopped in September 2021.

A total of 334 pwMS, were recruited, of whom 196 (58.7%) received their follow-up in a center with an MS-nurse and have been in contact with an MS-nurse at least once. The pwMS had a mean age of 44.5 ± 11.4 years and 68.9% of them were female. In centers with an MS-nurse, pwMS seemed younger, had a higher expanded disability status scale score (but not for the patient determined disease steps), were more frequently of male sex and had a higher level of education (Table 1).

Regarding our primary outcome, no clinical significant lower MSIS-29 total score (mean ranks: 169.9 vs. 157.8; Z = -1.114; p = .253) was found for pwMS followed in centers with an MS-nurse compared to without. MSIS-29 physical (mean ranks: 170.3 vs. 157.3; Z = -1.215; p = .224) and MSIS-29 psychological values (mean ranks: 169.6 vs. 159.7; Z = -0.927; p = .354) did not differ between groups either (Table 1 and supplementary, table S1). The MSIS-29 scores showed an asymmetrical distribution, therefore natural logarithmic transformations were applied. After this analysis, also no statistically significant mean differences between groups were found for the transformed MSIS-29 total (t(326) = 1.28; p = .200; 95%CI [−0.147; 0.031]), physical (t(326) = 1.28; p = .201; 95%CI[−0.162; 0.034]) and psychological (t(327) = 0.98; p = .330; 95%CI [−0.133; 0.045]) scores.

Since there were mean or median differences in the demographic variables (e.g. higher proportion of pwMS with a university degree in centers with a MS-nurse), we performed propensity score matching for the following clinically valuable variables: age, sex, employment, patient determined disease steps, expanded disability status scale, education and type of coping. After the propensity score matching, results remained similar, whereas no clinical differences were seen for the MSIS-29 (total) (mean ranks: 142.2 vs. 134.9; Z = −0.688; p = .492), nor were there differences for the MSIS-29 subscales, physical (mean ranks: 142.2 vs. 134.8; Z = −0.702; p = .483) and psychological (mean ranks: 142.0 vs. 137.0; Z = −0.477; p = .634).

Other quality of life measures, such as the SF36 and Eq. 5D, were also taken into account. These measures also showed no clinical meaningful differences between pwMS followed in centers with or without a MS-nurse (Table 2).

The results of the PROs showed no statistically significant lower proportions for anxiety (X²= 0.446, p = .800) and depression (X²= 1,772, p = .412) in pwMS who have access to an MS-nurse compared to pwMS without this access (Supplementary, table S2). T-tests for the coping inventory for stressful situations scores (CISS) measured no statistically significant different scores between pwMS followed in centers with or without a MS-nurse (task oriented t(322)= −0.65; p = .948, 95%CI[−2,513; 2,351]; emotional oriented t(322) = 0.917; p = .360, 95%CI[−1,367; 3,754]; avoidance oriented t(322) = 0.327; p = .744, 95%CI[−1,997; 2,793]) (Supplementary, table S3).

Since no clinical important differences were found between the two types of centers regarding anxiety, depression and coping, we continued to look at the data in general to find other explanatory relationships. For example, the Kruskal-Wallis test showed that a higher depressive (H(2) = 31.66, p < .001) and anxiety score (H(2) = 12.51, p = .002) was associated with a higher patient determined disease steps score. Intriguingly, ANOVA tests showed that pwMS who had borderline depression (mean difference 8.79; p < .001; 95%CI [4.29; 13.28;]) or depression scores (10.37; p < .001; 95%CI [5.77; 14,97]) and borderline anxiety (p < .001) or anxiety (p < .001) scores, have a statistically significantly higher score on emotional orientation than patients with normal depression or anxiety scores. Finally, the MSIS-29 total score was statistically significantly higher in pwMS who tested borderline abnormal or abnormal for depression (p < .001) and anxiety symptoms (p < .001) (Fig. 1).

Interestingly, a higher grade of MS-specific knowledge in pwMS followed in centers with a MS-nurse could be demonstrated (17.08(3.37) vs. 15.30(3.39), t(331) = 4.734, p < .001) and these results remained comparable after propensity score matching (17.07(3.24) vs. 15.7(3.26), t(280)−3.200, p = .002 95% CI [−2.22; −0.53]).

We performed comparisons for health care resource use and expenditures between groups with and without an MS-nurse. No statistically significant higher employment rate proportions were seen in centers with a MS-nurse compared to centers without (65.3% vs. 57.2; X²= 2.232, p = .135). Furthermore, no statistically significant differences were found in work change due to MS (15.8 vs. 11.7%, X²= 0.613, p = .536), income changes (57.9% vs. 87.5, X²= 2.220, p = .136), sick leave during the past three months (31.3% vs. 22.8%, X²= 1.736, p = .188), permanent sick leave (68.2% vs. 53.4% X²= 2.825, p = .093) and permanent sick leave due to MS (73.4% vs. 56.9%, X²= 3.688, p = .055). We found a higher proportion of hospital admissions (12.8% vs. 3.6%, X²= 8.261, p = .004) in centers with a MS-nurse, but no differences were found for admissions to a rehabilitation center (3.1% vs. 1.4%, X²= 0.900, p = .353) (Supplementary, table S4). No differences were found regarding follow-up by a general physician, doctor-specialist or other paramedics (Supplementary, table S5 and table S6). We investigated potential differences of several standard follow-up investigations (brain or spinal cord magnetic resonance imaging, ECG, blood analysis,.) but no proportional differences were shown except for evoked potentials, which were more frequently performed in centers without a MS-nurse (1.4% vs. 8.9%, X²= 7.499, p = .015) (Supplementary, table S7).

The data showed that a higher proportion of pwMS in centers with a MS-nurse have home adjustments (26% vs. 16.7%, X²= 4.108, p = .046) and are in need of health services (14.8% vs. 7.2%, X²= 4.475, p = .034). Yet, no proportional differences were reported regarding a home nurse (17.2% vs. 10%, X²= 0.300, p = .584), household help ( 69% vs. 88.9%, X²= 0.447, p = .504) or mobility services (20.7% vs. 20% X²= 0.002, p = .936) (Supplementary, table S9).

This exploratory, cross-sectional study aimed to assess the added value of an MS-nurse on HRQoL, health care resource use and health care expenditures, by comparing multiple PRO measures in centers with and centers without an MS-nurse.

The HRQoL in pwMS is known to be lower than in the general healthy population [11‐13]. Also, their perceived HRQoL has been reported to be lower compared to persons diagnosed with other chronic diseases such as epilepsy or diabetes [14]. The MSIS-29 questionnaire (total, physical and psychological) was used as primary outcome measure as it has been proven to be a valid and reliable measure for HRQoL and has the ability to assess MS specific domains [15]. Our data could not demonstrate clinically important differences nor associations with HRQoL (MSIS-29, Eq. 5D and SF36) between centers with or without the availability of a MS-nurse. Presumably, the MSIS-29 is more sensitive to disability discrimination and co-morbidities, such as anxiety and depression, than to assess the effects of differences in health care [16]. Also, the cross-sectional nature of our study, does not allow to assess changes over time in health-related quality of life, longitudinal follow-up may be more sensitive to detect differences between groups. Selection of pwMS based on disease duration or disability state, such as newly diagnosed pwMS or those with a long history of MS, rather than including all pwMS at once, next to predefined MS nurse task criteria, may be more efficient to address specific MS-nurse care and could be another way to demonstrate the added value of a MS-nurse.

Looking into the entire cohort, our data demonstrated lower HRQoL in pwMS who reported high scores for depression and anxiety. A previous meta-analysis showed that the risk of developing a depression (30.5%) or anxiety disorder (22.1%) is increased in MS. Importantly, the prevalence of, often undiagnosed, clinically significant depressive or anxiety symptoms is even higher compared to depression or anxiety disorders [17]. In our study, little over 20% showed a borderline depression or abnormal depression score, which is in line with the literature. Anxiety was also present in a similar proportion. Fast screening and individualized treatment of depression in pwMS has already shown to improve HRQoL [18]. Nurse-led stroke aftercare, for example, has already proven to decrease anxiety and emotional problems [19]. Long-term effects of an MS-nurse on depression and anxiety are potentially not assessed adequately here, as only one contact with a MS nurse was enough to include patients in this cross-sectional study.

Lower HRQoL was seen for pwMS who scored high on emotion-oriented coping. Generally, pwMS who adapt task-oriented instead of emotion-oriented coping strategy report a higher HRQoL and often experience a higher self-esteem [20, 21]. The latter type of coping, on the other hand, is associated with an increase of emotional distress, higher risk of depression and depersonalization [22]. Another study examined emotion-oriented coping strategies in non-depressed pwMS and showed that this subgroup more often adapts a passive coping posture against stressful situations because they may feel limited in their abilities to influence the disease course [23]. Emotion-oriented coping is also negatively associated to self-management. On the other hand, emotion-oriented coping is positively associated with the perceived MS impact [24]. Prior research findings allow to presume that psychometric variables are an even more salient correlate to health-related quality of life than objective clinical parameters [11]. These findings emphasize the importance of screening for illness perception and maladaptive coping strategies in order to improve self-management interventions in pwMS [25, 26]. Moreover improving self-management is considered as an important task for nurses in chronic care, yet the skills to provide such care are generally insufficient due to structural limitations such as education or time [26].

This study also evaluated the level of knowledge on MS in the study pwMS. The mean MS knowledge scores were higher in centers with a MS-nurse, even after propensity score matching where the educational level was balanced out between groups. The importance of education in healthcare has already been emphasized. For instance, one-on-one nurse-led education proved its benefits on HRQoL in patients with heart failure and epilepsy according to a systematic review and a specialist nurse intervention [27, 28]. The same systematic research also showed that nurse-led education could reduce hospitalization [28] and healthcare-costs [29]. Moreover, education might play a key role in improving autonomy and self-efficacy which could benefit therapy adherence [30]. This may indicate the importance of the educational role of an specialist MS-nurse [31] and further research investigating these parameters in MS is necessary.

There were between-group demographic differences concerning age, level of education, EDSS and disease duration. Indeed, the centers with a MS-nurse recruited a younger population with a higher expanded disability status scale score, level of education and need for home adjustments. These findings may be an indication of structural care differences, rather than the impact of a MS-nurse. For example, centers with an MS-nurse had more pwMS with a higher expanded disability status scale or patient determined disease steps scores in follow-up. On the other hand, healthcare expenditures were also included in this study but no substantial differences were seen regarding frequency of consultations or healthcare resource use. In other chronic diseases, there is evidence that involvement of a clinical nurse specialist or nurse practitioner as standard of care improves the quality of care regarding clinical outcomes and reduces the hospital length of stay [32‐34]. Previous cost-effectiveness studies showed mixed results, were context-dependent and often suffered from low methodological quality [33]. In the field of stroke and mental disorders, interventions and outcome measures were strictly defined and nurse-led care was able to reduce direct medical costs [35, 36]. Setting up a prospective study to evaluate the added value, including healthcare expenditures for an MS-nurse is necessary, but the design is complex as there are many aspects to consider such as study type, primary outcome, comparator, public payer or societal impact and time horizon (short- and long-term costs) [34].

This study was not able to show a higher HRQoL score in pwMS who have access to an MS-nurse compared to those who don’t. Other factors such as grade of disability, anxiety and depression, and coping strategies might play a more decisive role in the impact on HRQoL. However, there was a higher MS specific knowledge score in centers with an MS-nurse, potentially indicating the importance of an MS-nurse in educating pwMS. In order to strengthen the evidence of the role of an MS-nurse in the care pathway for pwMS, future research should focus on screening for maladaptive coping strategies and implementing MS-nurse-led specific education by performing longitudinal trials and a cost-effectiveness studies in general and university hospitals.